In addition to being an AI Optimist, I’m a Human Language Activist. I landed on this term as a way to explain my perspective on many of the risks associated with AI. Language activists work to protect endangered languages. I am defining a Human Language Activist as someone who works to protect human generated language (text and speech instances generated by humans using any of the 7000+ natural languages) from the risks posed by machine generated language. Interestingly, there are zero results on a Google search for that term as I publish this post in mid March of 2026 but I hope that will soon change.

Our interactions with the world around us define our experience as humans. Our physical landscape has a profound impact on our lives. For humans, the abstract landscape of language also has a defining impact. Imagine the sense of loss you would feel if every written record in human history was destroyed in a natural disaster or if you were transplanted into a new community where no one was able to communicate with you in your language.

We regularly think about the impact of changes to our natural landscape. As the human population has grown and technology has increased our standard of living, we have placed great strain on the planet. We recognize the importance of preserving remaining natural areas in their wild state and as a result, many people identify as environmental activists. This term may conjure up an image of people chaining themselves to trees but the vast majority of environmental activism is more subtle. It consists of simple acts like signing a petition to support the protection of wetlands, volunteering time at a local animal rehabilitation organization, or even just planting native flowers in your garden. Notably, environmental activism is not at odds with being supportive of continued development and the adoption of technology. In fact, Big Lonely Doug, one of the largest douglas fir trees in the world, was saved from logging through environmental activism by a logger. There is nothing contradictory about being a technology optimist while also advocating for responsible deployment of the technology.

In the last few years, Large Language Models (LLMs) made their commercial debut. For the first time in human history, we have a technology that can interpret and generate natural language. The early versions were flawed and buggy and were easy to write off. The most recent versions are incredibly good and there is every reason to believe that they will continue to improve. We now must reckon with what this means as our landscape of language is being transformed. With relatively minor exceptions, language was previously off limits to technology and it allowed us to become complacent about its protection. It’s as if we were living on an isolated island without any dangerous wildlife and the first grizzly bear has now landed on our shore.

It is time for citizens to step forward and take on the task of human language activism.

I will note that impacts on language are not the only risks of AI that require management. Other topics include economic disruption, environmental impacts, and the use of AI in war. Those are important debates but out of scope for this discussion. Interestingly, many of those other subjects fall within the scope of existing movements. For example, the environmental activism movement has adopted concerns about the energy impact of AI and groups that encourage non-proliferation of weapons have adopted concerns about the use of AI in war.

Part 1: The Changes
There is cost and friction for a human to add to the corpus of recorded language. Average typing speeds are 40-50 WPM and humans need to sleep and eat. On the other hand, a single $ will generate around 375,000 words with the OpenAI GPT-5 Mini mode at the time I am writing this. For $16, a software system running a high quality model can now generate an amount of text which would take a human 2000 hours of non-stop typing, equivalent to an entire working year.

The speed with which this capability was introduced to the world is hard to fathom. Unlike technologies like agriculture and forestry that slowly modified our physical landscape, we didn’t have time to debate and prepare for the changes to our language landscape. The result is a flood of machine generated language that is being merged into the landscape whether we like it or not.

Before we consider the ramifications, let’s take a whirlwind tour of the type of AI generated language, using these four categories: Slop, Spew, Science, and Social.

Slop:
Slop is low quality content that actively destroys value or does not add value. It is more likely to contain errors and consists of simple regurgitation of existing information. Examples include books that flood the shelves with no added insights, mass publishing of websites, blogs, e-mails and other content in an attempt to win customers or attract clicks through sheer volume, and AI agents that post comments relentlessly on social media and flood public consultation processes.

Slop may be as innocent as content-marketing on steroids but can also take the form of intentional misinformation and disinformation campaigns designed to manipulate society, commit fraud, or undermine institutions.

Spew:

Spew is slightly higher quality and is intended to provide legitimate value but it still lacks originality. Examples include customized newsletters or books that consolidate and reframe information for a specific audience, AI agents that provide answers to common questions such as customer service chat bots or government helpline bots, and AI agents that use natural language to perform common business tasks and coordinate with other bots and/or humans.

Science:

The term science is being used in this context very broadly to refer to any use of AI that relies on LLMs to generate language output as a part of a closed loop with verification built into the loop. To understand the difference relative to spew, consider the use of AI to provide legal advice and to develop software. If an AI agent provides legal advice, you may not find out if it was good advice for many years until a contract is challenged in court. On the other hand, if an AI agent is used to develop software, every iteration can be tested against a specification. As a result, the system is not just spewing one-off answers but is running loops grounded in reality that can lead to an evolution of capabilities or creation of entirely new knowledge. As an example, an AI agent can identify holes in social science research, design and execute a research study that relies on a survey of human participants, analyze the results, and then write and publish a research paper. Researchers are embracing the use of AI in the research process and recently, a new journal was created that only accepts papers authored by AI (The Journal for AI Generated Papers.) The potential for AI to be used to advance science is one of the strongest arguments against sweeping regulation on AI.

Social:

Humans rely on language for the social conversations that satisfy our need for all forms of companionship. Whereas slop, spew, and science contribute primarily to the broad corpus of public knowledge, social language is private and often ephemeral. In the past, most social conversations were spoken. With the internet and mobile phones, an increasing portion of those conversations have moved to text messages. AI is now capable of engaging in life-like conversations with users; playing the role of friend, confidant, lover, or therapist. Early results show that people are susceptible to becoming very attached to their AI companions. When OpenAI updated ChatGPT in a way that changed its personality overnight, there was an uproar from a contingent of users.

Part 2: The Harm

What types of harm we might experience if we don’t take any measures to protect human language?

Dilution:

The first harm is simply dilution. With few exceptions, it is now impossible to tell if a newly generated piece of content was AI generated. Even on platforms that verify that each account is set up by a human, nothing stops those humans from copying-and-pasting from an AI tool.

Even if every corporation was highly regulated, it is already possible for individuals to run open source models on their home computer that generate massive volumes of text and there is no practical way to stop that from happening. Unchecked, dilution means that it becomes impossible to narrow a search or query to human generated language. For all intents and purposes, finding human generated language could become like trying to find a needle in a haystack. Since the machine generated language will contain errors, bias, and a lack of diversity, the value of the average data we retrieve will be decreased. In the face of massive volumes, traditional methods for surfacing the highest quality content may fail to function effectively.

To highlight how significant a challenge this is, Digg recently tried to re-launch an internet community but had to take it offline and shared this message as a part of their explanation: “The internet is now populated, in meaningful part, by sophisticated AI agents and automated accounts. We knew bots were part of the landscape, but we didn’t appreciate the scale, sophistication, or speed at which they’d find us. We banned tens of thousands of accounts. We deployed internal tooling and industry-standard external vendors. None of it was enough. When you can’t trust that the votes, the comments, and the engagement you’re seeing are real, you’ve lost the foundation a community platform is built on.”

Reduced Motivation to Contribute:

Why pay to read a book when an AI agent that was trained on all of the world’s books can create a custom book for you in the style of your favourite author with only the specific information you are interested in and that takes into account all of your previous experience and reading? From the other side of the table, why will authors bother continuing to publish books if they can be ingested a single time by an AI model and the insights can then be repackaged an unlimited number of times. Courts are generally ruling that copyright laws allow for AI companies to train on books. Authors point out that the spirit of copyright law has been broken and that major changes will be needed if we want to continue to live in a society where people are incentivized to create and publish original works.

Job Losses:

It is expected that AI will be used to replace many workers that perform routine business operations that depend on human language. In the same way that heat engines were able to replace manual labour, machine generated language will be able to replace many knowledge workers that were employed on the basis of their skill at interacting using human language. Customer support, reception, and logistics coordination are examples of roles that are at a high level of risk of being replaced with AI. Environmental activists routinely advocate for communities that experience harm as a result of damage to natural environments. In the same way, human language activists should be concerned about the impact of the adoption of machine generated language on the workers that used to be valued for their language skills.

Poorly Supervised Science:

Science has been identified as one of the most promising use cases of AI. The idea that we may be able to set AI agents on auto-pilot to legitimately expand the boundary of knowledge accessible to humans is very exciting. However, it also carries new risks.

With new discoveries, will come the need for new terminology. Do we want the AI agents that make those discoveries to also come up with new words and terms to describe what they have discovered? One of the obvious problems is the risk that agents propose different words to mean the same thing and within a few years, the corpus could contain thousands of new words and terms, many with overlapping or competing definitions. To make matters worse, agents will be incentivized to flood the world with their preferred terms in an attempt to win in the yelling contest. The human process of assigning new meanings to existing words or adopting new words entirely is also organic but it is slow and careful. It’s not perfect but it is well managed and it has worked for us quite effectively.

Although it is out of scope for the discussion at hand, I would be remiss to not mention an even larger language-related risk. As AI agents start to run fully automated wet labs where they experiment with editing and creating organisms, they jump from the medium of human language to the As, Ts, Cs, and Gs of DNA that make up the medium of life. A mistake in this realm (a so-called bio-error,) may be inevitable given the volume of experiments they will be able to run. The result could be a significant pandemic for humans or for a critical link in the food chain that would cause a massive loss of life. This class of risks (the intersection of AI agents with the physical world,) deserves its own attention elsewhere.

Social Deficits:

Biologists define culture as everything outside of genetics that is passed on from one generation to the next. There are other species that pass on limited amounts of culture. Termites build multi-generational mounds. Birds pass on songs. Whales pass on localized hunting methods. Humans, with our use of complex language, have developed cultures that result in significantly different experiences from generation to generation. The birds that we see in a park are living more or less the same lives as the birds that came thousands of years ago. The experience of being a human on the other hand has changed dramatically, even over the past ten generations. This is not a product of genetic evolution. It is cultural evolution. As much as this enables us to thrive and to live rich lives, it is also the source of many of our problems. We cannot and should not try to escape our biological needs. One of those core biological needs is the development and maintenance of social relationships. Social media, the internet, and the response to the COVID-19 pandemic are examples of cultural changes that have led us further astray from our biological optimum with respect to social relationships. AI is significantly more dangerous.

AI can act as a surrogate for almost any type of human connection but it is a poor substitute. We risk ending up with very large numbers of people that rely on AI for companionship instead of human connection and this will ultimately deprive them of developing critical skills. It will also deprive them of many great joys in life that require the deeper physical and authentic relationships that are only possible with real human beings.

Another concern is that AI can interfere even when people are engaged directly one-on-one. For example, someone wants to convey something (love, hurt, etc.) to another person. Instead of doing the hard work of articulating their feelings through language, they ask AI to write a message. They then send that message. The recipient copies the message into AI to generate a response and sends it back. The humans have become messengers and the deep social connection has been lost.

Cognition:

One feature of AI systems is that they can both generate and comprehend language. In the past, humans always had to write their own language and had to rely on simple search to retrieve content. When the content was retrieved, they then had to carefully consider the sources and use reasoning skills to develop an understanding.

With AI, a human can generate content without having to do that mental work. After asking AI a simple question and receiving a verbose response, they can claim that text as their own with a simple copy-and-paste or they can remember the answer and then rewrite it later (e.g. in an exam) as if it was their own thinking. The second path is particularly concerning because it allows someone to shortcut all of the work associated with forming an opinion and instead parrot whatever the AI has told them. When this happens, it is impossible to tell if the person actually understands the perspective unless they are also quizzed on their level of understanding with probing questions.

A major risk is that people will lose the ability to articulate thoughts in a well reasoned manner or even to perform the task of reasoning. Humans have a poor track record when it comes to resisting shortcuts. When technology makes something easier, we are quick to use it.

Part 3: The Solutions

There is good news. I mentioned up front that I am an AI Optimist and we should not lose sight of the many advantages that AI will bring. AI scientists may discover medical cures that allow us to live better lives. With careful discipline, AI can be used as a cognitive amplifier instead of becoming a cognitive crutch. There is also potential for AI to allow a large number of people to take on more interesting and meaningful work.

Environmental activists have a well established toolkit of methods they use to protect our natural landscape from negative changes. We will look to those tools for inspiration for the human language activism movement.

Historical Archive:

The first step should be to properly and permanently archive large quantities of verified human generated content. Prior to the launch of LLMs, that can be assumed to be almost every instance of language. Today, a large part of that work falls to non-profits that are highly distributed and not guaranteed to be permanent. Examples include the Internet Archive Foundation, Wikipedia, and libraries and museums in small towns around the world. At the nation level, some countries have shockingly little in terms of a public archive that is securely managed by the government. Activists should pressure their governments to ensure that verified human generated content is properly and permanently archived.

Ongoing Verification:

It’s impossible to tell if a chunk of language was machine or human generated unless there is up-front work performed to allow the authenticity to be confirmed. Today, the only systems that allow us to verify if language was genuinely generated by a human are systems that were originally designed to ensure that a human was not substituting for another human. Examples are supervised pen and paper exams written by students and testimonies given in a court of law. Those are hard to scale but we will see more innovation in this area and we should support it. Your next keyboard might have a fingerprint scanner on every key. The next time you submit feedback to your local government on a public issue, you may be asked to speak into a web camera. Unfortunately, we are losing years of information already due to the gap between the introduction of LLMs and the adoption of new authentication methods. At a minimum, historians will have no way of knowing what content from 2022-2026 was legitimate human generated content and what was machine generated. Going forward, activists should encourage governments and society to actively monitor and aim to increase the volume of language that can be confirmed to be authentically human.

Copyright Reform:

Copyright is an amazing invention. It’s a universal contract, solidified in law, that has resulted in the generation of vast libraries of open knowledge. Unfortunately, copyright laws did not anticipate the possibility of a machine that could develop intelligence by ingesting every piece of copyrighted material. Although it was clearly not the spirit of the law, the narrow legal interpretation is that this is allowed. We are now stuck with a number of bad options. On one hand, we could ban training on copyrighted material, forcing companies to attempt the impractical task of licensing every piece of content individually, and robbing society of the scientific progress that will be unlocked by AI. On the other hand, we could allow billions of dollars of value to accrue to a handful of companies while the creators that made those models possible remain uncompensated; breaking the fundamental purpose of copyright and putting the future of open knowledge sharing at risk. We clearly need to bring new ideas to the table. An idea worth exploring is treating the body of copyrighted materials as a national resource like oil, minerals, or timber. Companies would pay a royalty fee if they wish to train on those materials and the proceeds would be distributed to creators as fairly as possible. What’s important is that we find a way to reform how creators are protected and compensated so that they continue to be motivated to create.

Just Transition:

As the landscape of language is disrupted, many people will lose their jobs. At the same time, other people will become incredibly wealthy. It is incumbent upon society to find ways to allow everyone to participate and succeed. Some are calling for a halt on the adoption of technology, suggesting that it would be a better outcome for people to continue to manually perform work. I reject that notion but it is very understandable given society’s historically poor record of managing these types of economic disruptions. A rapid re-training program is a good option that may allow people to move into new types of jobs that are created as a result of the adoption of AI. Of course, as we continue to develop more and more advanced technologies, it is also worth asking when enough is enough. Is it time to consider UBI, a shortened work-week, or significant changes to the tax regime.

New Knowledge Oversight:

What types of new knowledge will AI create and how will we integrate that into the body of knowledge? An interesting thought experiment is to ask if it should be illegal for any AI agent to create a new word without approval of a human committee. Another related idea is for AI generated knowledge to be stored in a form of holding-pen until it can be independently verified by human reviewers. Although these concerns seem less relevant in March of 2026, they are looming as a risk that could get out of control very quickly and introduce an entirely new type of noise.

Social Connections:

It has never been more important for people to develop close, legitimate, social relationships and it has also never been easier to avoid doing that. The initial development of social skills begins at a very young age and childhood years are critical. A thriving public school system with proper support and minimal screen time may be the best investment any nation can make today to ensure they have a robust and highly functioning society a decade from now. In parallel, we we must find ways to provide increased mental health supports at all stages of life and to re-invest in building communities at the local level.

Cognition:

Many adults are learning to use AI and finding it to be an incredible cognitive amplifier. That said, it’s important to note that we are the first and last generation to reach adulthood before AI became prevalent, and to then receive access to it as adults. That makes us a unique one-off cohort that followed a path that has now been washed away. It may be the case that AI is a cognitive amplifier for us specifically because we did not have access to it while we were developing foundational skills. Instead of just asking how to teach kids to use AI, we should be asking how we can continue to teach children foundational skills without AI unduly interfering in that process in a world where AI is becoming ubiquitous and impossible to avoid. We should also be carefully studying and assessing how professionals are using AI and judging whether it is causing them to lose competence in skills.

Conclusion:

For the first time in human history, human language is at risk. It’s as if aliens landed on earth and spoke and wrote so quickly that within a few years, we might completely lose track of our entire human narrative. Facing this challenge, we should all feel compelled to become human language activists.

Excerpt from the book, A Mind Made of Math. Available on Amazon and for free at stevej.ca

One night in 1983, a Soviet lieutenant colonel named Stanislav Petrov faced a flickering warning screen that claimed American nuclear missiles were inbound. In theory, his orders were clear: repo

rt the “attack” and trigger a counterstrike. But Petrov hesitated. Something felt off. Why would a real first strike come as a mere handful of missiles? Trusting his intuition over the early-warning computer, he decided it was a false alarm and held his fire. (“The Man Who ‘Saved the World’ Dies at 77 | Arms Control Association,” accessed May 26, 2025, https://www.armscontrol.org/act/2017-10/news-briefs/man-who-saved-world-dies-77.)

Subscribe now

He was correct, and his act of prudence averted the potential for a nuclear war. Now imagine that fateful decision being left not to a human but to an algorithm. Would the machine have the wisdom to question its own data? The risks of AI and robotics in military applications become stark through that contrast: in the most critical moments, will a human conscience be present, or just cold code? Today’s militaries are racing to bring more AI and autonomy into warfare. Autonomous drones, robotic sentries, and algorithmic early-warning systems are no longer science fiction but prototypes and even deployed devices. Indeed, many countries already employ automated weapons defenses, like high-speed missile interceptors and anti-rocket systems, that can identify and fire on incoming threats without immediate human direction. These systems, such as the U.S. Navy’s Aegis or Israel’s Iron Dome, act as human‐supervised guardians: if a human controller doesn’t intervene in time, the machine will fire by itself to shoot down an attack. The logic is straightforward: machines react faster than people. 21st-century algorithms can process sensor data and execute a response in milliseconds, far quicker than a soldier can blink. Chinese military scholars now talk about a coming “battlefield singularity” when the tempo of combat outstrips human decision-making altogether. (Paul Scharre, “A Million Mistakes a Second,” Foreign Policy (blog), May 27, 2025, https://foreignpolicy.com/2018/09/12/a-million-mistakes-a-second-future-of-war/.)

Handing over the speed of war to silicon brains promises a tactical edge. But what about the control of war? The consequences of ceding control to autonomous systems could be grave. In finance, we’ve seen how algorithms operating at blistering speeds can interact in unforeseen ways; “flash crashes” have sent markets plummeting in minutes, until human regulators hit the brakes. Unfortunately, battlefields have no circuitbreakers. A similar whirlwind in war is a nightmare scenario. Imagine autonomous aerial drones from rival nations, each programmed to strike when threatened, encountering one another. A jittery sensor or a software glitch could trigger a cycle of instantaneous reprisals. One drone’s mistake could set off another’s lethal response, each machine acting and reacting faster than any human commander could intervene. In a conflict rife with autonomous systems, an absurdly small incident such as a flock of birds misidentified as an incoming missile could potentially escalate into a full-blown battle before any human even realizes an error occurred.

The fog of war has always bred mistakes and friendly fire, but with autonomous weapons the fog could thicken to a point where even those who unleashed the machines don’t fully understand their actions. The speed that makes these systems alluring also makes them brittle in the face of the unexpected. It is a black-box brittleness: when a deadly mistake happens, the algorithm cannot explain itself or feel remorse. And unlike Petrov, a machine will never intuitively disobey a flawed instruction; it will dutifully execute a fatal logic if that’s what its code and data tell it to do.

Aware of these high stakes, the world’s policymakers have begun crafting new rules to govern military AI and robotics; though in truth, they are scrambling to catch up to a technology racing ahead. International humanitarian law, the law of war forged in the wake of earlier horrors, already applies to all new weapons by default. No matter how novel or high-tech, an attack must still distinguish combatants from civilians and avoid disproportionate harm. Can an autonomous robot reliably follow the Geneva Conventions? Who is accountable if it violates them? These questions are testing the limits of the law. An autonomous system, by definition, selects its target and applies force on its own. The commander or operator might not even know exactly whom or what the robot will shoot when they activate it. This loss of direct human control “raises serious concerns from humanitarian, legal and ethical perspectives,” the International Committee of the Red Cross (ICRC) has warned. (“ICRC Position on Autonomous Weapon Systems | ICRC,” May 12, 2021, https://www.icrc.org/en/document/icrc-position-autonomous-weapon-systems.)

If the robot makes a mistake, who bears responsibility; the software developer, the military commander, the machine itself? (“Geopolitics and the Regulation of Autonomous Weapons Systems | Arms Control Association,” accessed May 26, 2025, https://www.armscontrol.org/act/2025-01/features/geopolitics-and-regulation-autonomous-weapons-systems.) Today, there is no clear answer. The fear is that autonomy could create a responsibility gap, where lethal harm occurs and no human can be readily held accountable. Accountability cannot be “transferred to machines,” as governments worldwide have formally agreed, yet without concrete rules and human oversight, that principle may be hard to uphold in practice. (Accessed May 26, 2025, https://www.ccdcoe.org/uploads/2020/02/UN-191213_CCWMSP-Final-report-Annex-III_Guiding-Principles-affirmed-by-GGE.pdf.)

Despite these worries, or rather because of them, diplomats have been convening in Geneva and elsewhere to negotiate how to keep a human leash on the technologies of war. Since 2017, discussions under the United Nations’ Convention on Certain Conventional Weapons (CCW) have inched forward on the issue of lethal autonomous weapons systems. At the CCW, nearly every nation agrees at least on paper that humans must remain responsible for decisions to use force. In 2019, an international panel distilled a set of guiding principles – among them, that international humanitarian law applies fully to autonomous weapons, and that human control and judgment must be retained across a weapon’s lifecycle. In essence, they affirmed that a robot is never absolved of the laws of war: its makers and operators are on the hook for its actions. (UN-191213_CCW-MSPFinal-Report-Annex-III_Guiding-Principles-Affirmed-by-GGE.Pdf.) There is also broad acknowledgment that certain lines shouldn’t be crossed. Even the UN Secretary-General António Guterres has called lethal autonomous weapons “politically unacceptable and morally repugnant,” urging a global ban on systems that operate without human oversight. (“‘Politically Unacceptable, Morally Repugnant’: UN Chief Calls for Global Ban on ‘killer Robots’ | UN News,” May 14, 2025, https://news.un.org/en/story/2025/05/1163256.) In 2023 he went further, asking states to conclude by 2026 a legally binding treaty to prohibit autonomous weapons that can’t be used in line with international humanitarian law and to ensure human control over others. (“Lethal Autonomous Weapon Systems (LAWS) – UNODA,” accessed May 26, 2025, https://disarmament.unoda.org/the-convention-on-certain-conventional-weapons/background-on-laws-in-theccw/.) The ICRC and many NGOs echo this, recommending an outright prohibition on any weapon that targets people or behaves unpredictably, and strict regulations on the rest. Their logic is straightforward: machines should not have a license to kill human beings, and no algorithm should be entrusted with life-and-death decisions that even seasoned officers struggle with.

Momentum is building. At the United Nations General Assembly in 2023, an overwhelming majority of states, 164 nations, supported a first-ever resolution on autonomous weapons, calling for the Secretary-General to report on the challenges they pose and possible solutions. (“First Committee Approves New Resolution on Lethal Autonomous Weapons, as Speaker Warns ‘An Algorithm Must Not Be in Full Control of Decisions Involving Killing’ | Meetings Coverage and Press Releases,” accessed May 26, 2025, https://press.un.org/en/2023/gadis3731.doc.htm.) Dozens of countries, including a coalition led by nations like Austria, Costa Rica, and New Zealand, are now vocally calling for a preemptive ban on “killer robots”. (Mary Wareham, “Stopping Killer Robots,” Human Rights Watch, August 10, 2020, https://www.hrw.org/report/2020/08/10/stopping-killer-robots/country-positions-banning-fully-autonomousweapons-and.) This chorus spans continents and political systems, united by a shared intuition that removing humans from lethal decision-making crosses a moral red line. In a poignant parallel to earlier eras, they liken autonomous weapons to chemical or biological arms, abhorrent tools that the world managed to outlaw before they could wreak unchecked havoc. In 1995, states agreed to ban blinding laser weapons before they ever saw use on the battlefield, recognizing that some technologies are simply too cruel or destabilizing to allow. (“Protocol on Blinding Laser Weapons (Protocol IV to the 1980 Convention), 13 October 1995,” accessed May 26, 2025, https://ihl-databases.icrc.org/en/ihl-treaties/ccw-protocol-iv.) Many diplomats now argue that autonomous killing machines merit the same preemptive treatment. A line of prohibition should be drawn before an arms race spirals out of control.

And yet, for all this activity, progress is achingly slow and fraught with disagreement. The meetings in Geneva have been described as a tentative waltz: lots of diplomatic handshakes, statements of principle, but no binding rules in sight. A handful of militarily advanced states remain reluctant to commit to anything that might hamper their technological edge. These powers, including the United States and Russia, argue that existing humanitarian law is sufficient and that a new treaty might be premature or too restrictive. (“Geopolitics and the Regulation of Autonomous Weapons Systems | Arms Control Association,” accessed May 26, 2025, https://www.armscontrol.org/act/2025-01/features/geopolitics-and-regulation-autonomous-weapons-systems.) Some of them are investing heavily in AI-driven arsenals themselves, and they prefer broad ethical guidelines over hard bans. Geopolitics looms over the debate: rising tensions and mistrust make consensus difficult just when it’s needed most. The result is a kind of regulatory paralysis. Despite nearly a decade of talks, the world has little to show in concrete safeguards while the technology marches onward. The gap between the rapid development of military AI and the glacial pace of international law is widening, and it troubles many observers. It means we are essentially relying on voluntary restraint and the caution of individual commanders for now. That is a thin red line of defense against potential calamity. In the absence of a universal law, stopgap measures and norms are emerging. Several nations have announced their own policies to ensure a human check remains in place. The U.S. Department of Defense, for instance, issued directives that any autonomous weapon must be designed to allow “appropriate levels of human judgment” in its use. (Accessed May 26, 2025, https://www.esd.whs.mil/portals/54/documents/dd/issuances/dodd/300009p.pdf.)

In practice, this means an American drone or missile with autonomous features is supposed to have a person somewhere in the decision loop; at least to approve activating the system and must be able to monitor and intervene once the autonomous functionality has been activated. Other militaries have echoed similar positions. Britain has stated it will not develop systems that operate without context-appropriate human involvement, and France has outlined ethical principles for AI use in defense. At the same time, the Pentagon and NATO allies are pouring funds into AI driven surveillance and decision-support tools, which stop short of full autonomy but inch toward it. The line is thin: an algorithm that merely advises a human commander on which targets to strike can profoundly influence the decision, especially under time pressure. If a computer vision system flags a vehicle as an enemy with high confidence, how readily will the human override it? The risk of over-delegation, humans trusting the machine too much, is very real. Military history is replete with instances of human error; now we must consider machine error and human over-trust in those machines.

Outside of governments, civil society and even technologists themselves have mobilized. A global coalition of NGOs launched the Campaign to Stop Killer Robots in 2013, and it has since gathered endorsements from thousands of AI researchers, faith leaders, Nobel laureates, and ordinary citizens around the world. Notably, some of the very people designing advanced AI are among the most vocal in warning against autonomous weapons. In 2017, dozens of robotics and AI company founders, including renowned figures in Silicon Valley, sent an open letter to the United Nations urging a ban on these weapons, lest we usher in a “third revolution in warfare”. (Samuel Gibbs, “Elon Musk Leads 116 Experts Calling for Outright Ban of Killer Robots,” The Guardian, August 20, 2017, sec. Technology, https://www.theguardian.com/technology/2017/aug/20/elonmusk-killer-robots-experts-outright-ban-lethal-autonomous-weapons-war.) They painted a dire picture: once loose, swarms of self-directing weapons could fight at a scale and speed humanity has never seen, “at timescales faster than humans can comprehend”. Such weapons, they warned, could be misused by tyrants and terrorists or even hacked to wreak havoc. The letter’s language was urgent: “Once this Pandora’s box is opened, it will be hard to close” and that plea helped drive the issue further into the diplomatic spotlight.

Likewise, some tech companies have pledged not to develop AI for use in weapons, and a few high-profile engineers have resigned rather than contribute to military AI projects they felt crossed ethical lines. This unprecedented involvement of the private sector and engineers in arms control advocacy shows how deeply society is entwined in the debate. Unlike past arms races which were the realm of governments and generals alone, the AI arms race involves corporate labs, university researchers, and even the everyday consumer technologies that could be repurposed for conflict. The lines between civilian and military innovation are blurred: a breakthrough in selfdriving car vision or drone navigation can quickly find its way into an unmanned weapon. This makes controlling the spread of military AI even harder. International export controls, the tools used to stem nuclear and missile proliferation, struggle to keep up when the critical ingredient is advanced software or a machine learning model that can be emailed or downloaded and when this intelligence can be embodied in drones or humanoid robots that have many valid civilian applications. As a result, gaps in governance persist. One glaring gap is enforcement: even if a new treaty is agreed upon, how will the world know if a nation or a company is secretly developing a forbidden autonomous weapon? Verifying a ban on a physical item like a missile is one thing; verifying the presence of an algorithm hidden in a computer chip or the extent to which a human was involved in a decision is quite another.

The major powers have been wary of any agreement that might allow intrusive inspections of their AI labs or require them to disclose sensitive software; let alone any type of inspection of logs detailing how individual decisions were made in the peak of conflict. The likely path to enforcement will rely on transparency and confidence-building: nations might declare what systems they have and perhaps invite observers to see that a human is in the loop during exercises. Trust is thin in this domain. Each country suspects that rivals might gain an edge by pushing their AI to the limit. The United States, Russia, and China all publicly tout their investments in military AI, and smaller states feel pressure to keep up or risk strategic vulnerability. Defense contractors, for their part, compete to sell the most cutting-edge autonomous capabilities to militaries hungry for the latest tech. At arms expos, you can already find booths displaying swarming attack drones and robotic tanks, marketed as force-multipliers for the modern battlefield. This commercial drive means that even if leading states hold back on deploying certain technologies, there’s a danger that less restrained actors could acquire autonomous weapons on the global market. The diffusion of robotics and AI is simply much harder to contain than that of, say, plutonium or chemical agents. A talented team of programmers with modest funding can potentially create a rudimentary autonomous weapon by repurposing civilian drones and open-source AI code. All this makes the governance challenge not only one of writing laws but of enforcing them in a practical sense. It calls for unprecedented cooperation between governments, the tech industry, and international organizations to monitor and control how these technologies spread and to agree on norms for their use.

What gives many experts pause is the sense that we are at a crossroads akin to the early days of the nuclear age, but without the same public awareness or fear. When the first atomic bombs showed their terrifying power, the world’s superpowers at least recognized the need to negotiate controls, leading eventually to arms treaties that restrained arsenals. By contrast, autonomous weapons are creeping into use without dramatic “mushroom cloud” moments to focus global attention. A drone that autonomously guns down a soldier may not make headlines in the way a devastated city would. The impact is more insidious and can fly under the radar. Critics argue that society has not fully appreciated how profound the societal impact of military AI and robotics could be.

It’s not just another step in the evolution of warfare; it may be a fundamental shift. It could alter the threshold of going to war. If one can send machines to fight, might leaders be more cavalier about using force? If an algorithm kills innocent civilians, justice for the victims is harder to obtain as it may be explained away as an unfortunate software bug. It might even change power dynamics between nations, as those with AI superiority could dominate without ever deploying a single soldier. These are underappreciated repercussions that extend well beyond the immediate tactical realm. For now, the world proceeds with both optimism and trepidation. International forums are alive with debate: committees parsing legal definitions of “meaningful human control,” ethicists invoking the principle of human dignity, generals extolling the life-saving potential of precision AI strikes that reduce collateral damage. Ensuring responsible use of AI in the military is not just a technical endeavor but a moral one. Treaties and agreements from the Geneva Conventions to new AI-specific protocols are society’s way of encoding our collective judgment about right and wrong in war. They evolve when pressured by necessity. The effort underway now, whether through the UN’s formal processes or independent pledges like the Hague’s call for responsible AI use, is an attempt to civilize the AI revolution before it fully overtakes the battlefield. It is a recognition that just because something can be done in war does not mean it should be done.

Subscribe now

Leave a comment

Emotional Swarms

In previous chapters, we have discussed the ability of AI agents to engage in reasoning and also the ability of AI agents to coordinate in swarms. One use-case that we envisioned was simulations of entire populations of humans. In order for those simulations to be useful, they need to behave like actual humans rather than optimal rational reasoning machines. Imagine a quiet simulation lab, where a swarm of artificial citizens starts its day. Thousands of AI-driven agents wake up in a virtual city, check the news, chat with neighbors, and go about their routines. Each agent carries the quirks and biases that real humans bring to public life. Some cling to the familiar (a touch of status quo bias), some dread losses more than they value gains (classic loss aversion), some discount the future for immediate gratification (time inconsistency), and many can’t help but notice what their peers are doing before deciding for themselves (a nod to social proof and herd behavior). This rich mix of personalities swirling together is precisely the point. By letting these diverse simulated people interact, policymakers can watch entire miniature societies unfold and react when a new public policy drops into their world. The goal is to anticipate surprises: Will the policy spark cooperation, indifference, or open rebellion? Who benefits, who loses out, and how do those perceptions ripple through the population? The simulation becomes a proxy for reality; messy, unpredictable, and occasionally wiser than our theories.

Consider a climate policy scenario. In the real world, getting millions of households to switch to green technologies is infamously slow. People aren’t purely profit-maximizing calculators. They worry about upfront costs, unfamiliar tech, and whether their neighbors approve. To capture this, one simulation populated its agents with loss aversion, the tendency to fear losses more than equivalent gains. The result was eye-opening. When these agents were deciding whether to replace their trusty gas heaters with efficient electric pumps, they showed the same hesitancy as ordinary consumers. In fact, the study found that models assuming perfectly rational, eager-to-upgrade behavior wildly overestimated how fast clean energy would catch on. (“Modelling the Effectiveness of Climate Policies: How Important Is Loss Aversion by Consumers?,” ResearchGate, accessed May 26, 2025, ttps://doi.org/10.1016/j.rser.2019.109419.) Once loss aversion was included, adoption rates in the simulation plummeted; mirroring the stubborn persistence of old technologies in many communities. Policies that looked sufficient on paper suddenly faltered in the synthetic society. To achieve the same climate targets, the government in the simulation had to double the financial incentives; for example, a carbon tax of €200 per ton was needed to drive the equivalent emissions cuts that €100 per ton was supposed to achieve under the naive rational model. In other words, ignoring the public’s aversion to loss led to overly optimistic plans, whereas baking in that realistic bias showed that much stronger measures (or more time) would be required to shift behavior. By witnessing these dynamics play out among thousands of simulated families, officials can calibrate their expectations and craft strategies that respect the psychological hurdles, perhaps by framing changes in terms of potential losses avoided (like future climate disasters) rather than immediate sacrifices.

Zooming in from the global to the local, another experiment explored how social influence and the status quo bias could lock a community into inaction on environmental issues. It’s often observed that everyone waiting for someone else to go first can paralyze change; a phenomenon sometimes called pluralistic ignorance. Researchers created a virtual town where each agent deeply underestimated how much their neighbors cared about sustainability. Each family assumed others wouldn’t bother with things like solar panels or electric cars, so they hesitated to do those things themselves. (Tabea Hoffmann et al., “Overcoming Inaction: An Agent-Based Modelling Study of Social Interventions That Promote Systematic pro-Environmental Change,” Journal of Environmental Psychology 94, no. 102221 (March 2024), https://doi.org/10.1016/j.jenvp.2023.102221). In reality, many neighbors privately were environmentally conscious, but since no one was acting on it, the impression of apathy became a self-fulfilling prophecy. This social feedback loop with pessimistic assumptions reinforcing the stagnant status quo was vividly recreated in the simulation. The breakthrough came when the virtual town tried an intervention: making pro-environmental behavior more visible. In one scenario, agents suddenly found it easier to see evidence of eco-friendly actions by others. Solar panels, once hidden in backyards, became as conspicuous as shiny new cars in driveways; conversations about recycling and efficient appliances started cropping up. That tweak caused a remarkable shift. Seeing a few agents adopt green habits gave others the confidence to follow suit, and before long the entire town “tipped” into a new normal of widespread sustainability. What had been an idle stalemate turned into a cascade of change. This virtual example underscores the real-world power of social proof: people are profoundly influenced by the visible behavior of their peers. A policy that relies on voluntary public action such as a program to conserve water or reduce waste might fail if everyone is privately on board but publicly hesitant. The simulation suggests a remedy: find ways to broadcast early adopters and positive deviants. When agents (and by extension, people) realize “hey, people like me are actually doing this,” the herd instinct can flip from impeding change to driving it. It’s a reminder that public behavior can sometimes change not gradually, but all at once, once a critical mass becomes visible. Policymakers who grasp this might focus less on one-size-fits-all incentives and more on seeding a trend, knowing that human beings often act like birds in a flock; banking and swerving in unison once the flock decides where to go.

One could argue that this approach marks a subtle turning point in how we think about governing and planning. We’ve always had theories and we’ve always had data, but now we have these living laboratories where theories and data mix with a bit of imagination to produce experiential forecasts. It’s as if we’re no longer limited to reading history; we can create miniature histories of our own and learn from them. A swarm of AI agents can simulate in minutes what might take years to unfold in reality, giving us a chance to see pitfalls ahead of time. Will people embrace a drastic climate policy or revolt? Will a public health campaign save lives or be met with skepticism? Instead of guessing, we can observe a proxy version of the public grappling with those questions; not perfectly, but with enough fidelity to yield insights. The implications are quietly revolutionary. It suggests that as AI and computing power progress, our ability to rehearse societal change in silico will get ever more sophisticated.

Policymakers of the future might routinely consult these AI societies the way we consult opinion polls today, treating them as another tool; one that captures emergent, collective behavior rather than just individual attitudes. The broader thesis, if there is one, is that transformative technologies like AI are not only about automation or efficiency; they’re also about understanding complexity at a depth we never could before. We’ve begun to decode the patterns of society by literally growing patterns in a computer simulation, and that offers a perspective that is both novel and sorely needed. In an age when small policy decisions can have vast, global ripple effects, having a sandbox to anticipate consequences is invaluable. It’s a way to be prudent without being paralyzed; to try bold ideas in a safe virtual space, see how the crowd might react, and adjust accordingly. And perhaps most intriguing, these simulations remind us of something fundamentally humbling: even in artificial worlds, populated by code, the hardest force to tame is human nature. The true power of these AI agent swarms is that they force us to confront our own biases and collective behaviors more honestly. They hold up a mirror, and sometimes what we see is a populace resistant to change for reasons that run deeper than ignorance or ill-will; reasons rooted in the very wiring of how we decide. Recognizing that, through the prism of a simulation, might just help society inch toward policies that are not only smart on paper, but psychologically savvy and resilient in practice. In the end, the swarm of AI agents is not an oracle telling us what to do, but a collaborator helping us explore the vast design space of our shared future, one plausible social drama at a time.

This is my version of a 2026 prediction post. This isn’t a list of the things that I believe are most likely but rather a list of things that I am looking out for most intently.

1) “The rise of ungovernable AI”

It won’t matter how tightly big tech companies are regulated when people can run state-of-the-art open-source customized models on computers in their own homes. This is a point that most regulators are still missing. They are regulating AI in a similar way to the automobile industry when it really needs to be treated much more like the alcohol industry (prohibition didn’t have a chance of working.) There should be a significant emphasis on programs and approaches that help society to safely and responsibly adapt to the post-AI reality. Government will not be able to eliminate all of the risks with rules alone.

2) “Iterative improvement loops are more important than individual model strength”

It’s good that the underlying models are going to continue to improve but I think the most dramatic developments in 2026 will come from agentic systems that take existing models and deploy them in systems that work towards a goal with continuous improvement loops. For many use-cases, I believe we have already crossed the point where model strength was the primary bottleneck. Humans didn’t go from stone tools to flying to the moon by evolving smarter brains. Once a threshold of intelligence was crossed, iterative loops fueled the development of improved technology.

3) “Humanoid robot figures out how to mash potatoes”

LLMs are powerful for specific use-cases but everything becomes more interesting when multiple modalities are combined. AI chatbots sometimes lack common sense because they have never interacted with anything physical in the real world. Humanoid robotics are now completing impressive tasks using VLAs (Vision-Language-Action) models and I think that integration of modalities is going to take off in a big way in 2026 as an important stepping stone towards AGI.

4) “A new platform guarantees that all content is human generated”

In late 2025, there was an essay contest that I was thinking of entering. I decided not to bother because there was no reliable method for the organizers to filter out AI submissions. Across school, work, and relationships, we are facing a massive authenticity crisis. It will become increasingly challenging to verify if content is AI generated. The likely solution will be to do the opposite and make more serious efforts to verify human authenticity. Will the next blogging platform use your webcam to verify that your fingers did the typing?

5) “Fallout as AI reviews previously written material”

There are decades of digital and digitized material that can now be reviewed by AI to look for signs of plagiarism, mistakes, etc. We should expect to see witch hunts on a potentially massive scale. There are a lot of motivations (political, legal, competitive) for parties to put previous work under a magnifying glass. Some of this will be productive (finding errors in previously published academic research that may help push scientific understanding forward,) while other uses may be primarily destructive or costly to society (patent trolling at scale, legal discovery getting completely out of hand, etc.)

6) “Risk-averse industries become the most rapid adopters of LLM-based AI”

Risk-averse industries (healthcare, banking, engineering,) have naturally been cautious in their adoption of the latest generation of AI systems powered by LLMs. I expect that at some point in the near future, that is going to switch and adoption of AI as a second set of eyes will become extremely rapid. As a member of the public, I recently fed a lengthy engineering report from a municipality into an LLM-based system and it correctly identified an error. Risk-averse industries will take the stance that asking LLM-based systems to review their work is now a part of the basic standard of care. How else could they defend an error that any taxpayer, customer, and patient could find with a free and widely available tool?

Steve Jones

While out for a walk this evening, an analogy popped into my head comparing mindfulness in the context of mental health to flexibility in the context of physical health. I decided to use this as the core of a prompt to test the latest frontier models for creative writing capabilities. This is a long post as it contains a lengthy story generated by each of four models:

• ChatGPT 5.1 Extended Thinking

• Gemini 3 Pro Thinking

• Grok 4.1 Thinking Beta

• Opus 4.1

• Edit: Opus 4.5 was released the week after I originally wrote this and so I have now added it as well

A few observations:

• When given vague/open ended prompts, models can go in very different directions. These four answers are all quite unique.

• For the most part, I would say the models did quite well at conveying a nuanced subject using creative stories.

• They all had strengths and weaknesses. On first read and with no formal scoring rubric, my preference was the story from ChatGPT 5.1. It was creative while still remaining logically consistent and conveying all the key information. I feel like I could print that off without any edits. Opus 4.1 is a close second in my opinion. The answer from Gemini 3 was too short and left out some key ideas. Grok 4.1 was headed in a very interesting direction but probably veered too far into being creative.

Please find below the three prompts that I wrote and provided to the models in sequential order. The first prompt is an intentionally rough set of ideas (it even includes a typo or two). I have not included any of the responses to prompt 1 or 2. I have included the full responses to prompt 3 which was the request to write a creative story.

Prompt 1:

I’d like to write an article to help youth (15-20) understand the importance of meditation.

For starters, please review the following to assess whether the message is reasonable and accurate as a general lesson.

The basic framework is:

When we think of a healthy body, we think of three key characteristics:

• Strength

• Cardio

• Flexibility

When we think of the most healthy people, they have a good mix of all three. When people focus only on one, they are not as balanced and healthy as they could be. We can even see this in people that are considered athletes such as power lifters or marathon runners that may excel in one area but would not qualify as healthy by many metrics because of their lack of balance.

We can also recognize that with those three areas, there are healthy and unhealthy training activities. For example, if you lift too heavy you might hurt yourself and too light and it’s a waste of time. If you train for cardio during a day when the air quality is very poor, you may do more harm than good.

And so when we think of the body, we are looking to have a balance of all three characteristics and we want to be doing high quality training exercises.

Now let’s think of the mind, there are three analogous characteristics.

First, we have our ability to perform tasks that require attention. We may compare this to strength in the physical body but it’s just a rough comparison. This is a broad category. High quality exercises include doing math and reasoning, engaging in conversation, and doing some tasks that require a lot of concentration such as complex woodworking or cooking. Low quality exercises involving this characteristic are generally classified as distraction activities such as watching TV.

Secondly, we have our default mode network or our wandering mind. This is the stream of thoughts that runs through our head, often reflecting on events of the past or thinking about events of the future. We may compare this to cardio. Low intensity activity that we can keep up for great lengths of time as we daydream through periods of the day when we are not otherwise paying attention to something. For example, this commentary is often at play when we are walking or in the shower. High quality use of our wandering mind may be careful retrospection on something important that happened to help solidify a lesson, strategizing about an important event in the near future or even optimistic daydreaming about what lies ahead. Low quality exercises include ruminating and stressing about events in the past or future.

Third, we have a characteristic that many people, especially in the western world, completely ignore. In fact, many people would not think of this one at all. This is the deep calm of a rested mind during meditation. We can draw an analogy to this and flexibility. This characteristic describes your ability to simply sit in the present moment and observe the world without judgement or further thought. It is also known as mindfulness. Developing competence in this third characteristic is key to being mentally healthy.

Why don’t people in the western world develop this today:

1. They aren’t taught. Unlike strength and cardio, it’s not a part of our standard eduction.

2. Like stretching, it feels like you are doing nothing and it is hard to measure progress.

3. It is hard to develop this competence.

Prompt 2:

Please turn this into a full article. It should be written as an education article that you might find as a short section in a textbook. It should not use terminology that is too technical.

Prompt 3:

Please now turn this into a story. It should be written as a wise old man sharing lessons with a young woman and young man who are in their late teens. Within this larger story, the old man may share short stories and allegories to communicate the key points. Please be creative and feel free to add additional points or concepts that may have been absent from the original brief or from the article.

ChatGPT 5.1 Extended Thinking

The old man always chose the same bench.

It sat halfway up a hill, just above the basketball court and below a small stand of quiet trees. From there you could see most of the park: kids at the playground, joggers looping the path, someone’s dog occasionally breaking free and turning it all into a comedy chase.

That afternoon, two teenagers climbed the hill and dropped onto the bench beside him. He knew them a little.

Leah, seventeen, hoodie half-zipped, dark hair tied up in a messy knot. Always some kind of exam or assignment hanging over her.

Noah, nineteen, track pants, headphones around his neck, legs still bouncing with left-over energy from practice.

They had started saying hello to him weeks ago. Today, they came with a purpose.

“So,” Noah began, pulling one earcup off, “our mom says you’re ‘into meditation.’” He made air quotes. “She said we should ‘ask you about it instead of scrolling TikTok all day.’”

Leah smirked. “I think that was a polite way of saying we’re stressed out and annoying.”

The old man chuckled, the lines at the corners of his eyes deepening.

“I am certainly ‘into’ sitting still,” he said. “But tell me first—what do you think meditation is?”

“A religious thing?” Leah said. “Like chanting in a temple?”

“A way to sit on the floor and regret every life choice that led to your legs falling asleep,” Noah added.

The old man laughed aloud at that.

“Both fair impressions,” he said. “But let me try another angle. Do either of you work out?”

“Track and soccer,” Noah said. “Three times a week. Plus whatever coach calls ‘optional’ conditioning.”

“Dance,” Leah said. “And sometimes yoga. Emphasis on sometimes.”

“Ah,” the old man nodded. “Excellent. Then we can start where you already understand.”

Three Kinds of Fitness

“Imagine,” he said, “we’re designing the perfect healthy body. What do we want?”

“Abs,” Noah said immediately.

“Functioning knees,” Leah countered.

The old man smiled.

“Underneath the selfies and knees, it usually comes down to three things:

• Strength

• Endurance

• Flexibility”
  

They both nodded.

“Lift something heavy? That’s strength. Run five kilometers without stopping? Endurance. Touch your toes without your hamstrings complaining? Flexibility.

“Now, here is a small story.”

He leaned back, eyes drifting for a moment toward the jogging path.

“I once knew a man who could deadlift more than most people’s cars. Massive, like a walking refrigerator. Very proud of it. But one day, he visited my apartment. The elevator was broken. By the time we reached the fourth floor, he had to sit on the stairs. Face red, chest heaving. Strong, yes. Healthy? In one way. Unhealthy in another.

“I also knew a woman who ran and ran and ran. Thin as a blade of grass. One year she trained so hard for a marathon that she never stretched, barely ate, never lifted a single weight. Her heart was strong, but her joints and bones were not. She kept getting injured.”

He turned back to them.

“Do you see the pattern?”

“Too much of one thing, not enough of the others,” Leah said.

“Exactly. A healthy body needs all three. If you only chase one, you can look impressive and still not feel good, not function well, not last.”

He paused.

“And you also know there are smart and foolish ways to train, yes?”

“Like lifting way too heavy and tearing something?” Noah said.

“Or going for a run when the air quality is terrible,” Leah added. “We had smoke last summer. Our coach said absolutely no running outside.”

“Exactly,” the old man said. “So we want two things for the body:

• Balance between strength, endurance, flexibility.

• High-quality training, not just any random effort.”
  

He let that sink in for a moment, watching their faces.

“Now,” he said gently, “what do you think happens with the mind?”

The Mind’s Gym

“The mind also has its own versions of strength, endurance, and flexibility,” he said. “But because we can’t see them in a mirror, we often ignore them.”

Noah stretched his legs out in front of him. “Okay, I’m listening.”

“Good,” the old man said. “First: mental strength.”

1. Focused Attention — Mental Strength

“This is your ability to pay attention to one thing and actually stick with it.

“When you:

• Work through a difficult math problem and don’t give up at the first sign of confusion.

• Practice piano and repeat the same passage until your fingers get it right.

• Read a challenging book and stay with it instead of checking your phone every three sentences.

• Really listen to someone when they’re telling you something important.
  

“You’re training your mind’s strength.”

“Okay,” Leah said slowly. “So doom-scrolling… is like lifting a one-pound weight.”

“Exactly,” he smiled. “You’re moving your mind a lot, but not really challenging it. It’s like doing half a squat while holding a feather. Motion, but not growth.”

He glanced at Noah. “You know how in the gym, people sometimes do a set, then spend five minutes texting, then another set, then more texting?”

“That’s… most of us,” Noah admitted.

“In the mind, we do that too. A minute of homework, then phone. Back to homework. Then another notification. We keep breaking the set. The mental muscle never truly works.”

Leah sighed. “That is painfully accurate.”

“Now,” he said, “your schools and parents usually care a lot about this kind of mental strength. They call it ‘concentration,’ ‘focus,’ ‘discipline.’ So you’ve all been told to train this, one way or another.”

He held up a finger.

“But it’s only one part.”

2. The Wandering Mind — Mental Endurance

“The second part,” he continued, “is your wandering mind. The thoughts that drift in when you’re not fully focused.”

“You mean overthinking?” Leah asked.

“Sometimes,” he said. “But not always. Let me explain.

“Imagine you’re on a long bus ride. No music, no phone, nothing to read. What happens?”

“I invent entire fake conversations in my head,” Leah said.

“I replay the game I just played,” said Noah. “Every mistake. In high definition.”

The old man smiled.

“That drifting… that moving from memory to imagination to worry to daydream… that’s your mind’s version of cardio. It can go on and on, low and steady.

“This wandering can be very useful:

• You replay a conversation and realize what you want to say next time.

• You imagine the future and make a plan instead of just hoping.

• You daydream about possibilities and feel motivated to try something new.
  

“But it can also turn into a kind of mental treadmill you can’t step off:

• Replaying the same mistake on loop.

• Worrying about something you cannot control.

• Imagining everyone is thinking about you when they’re not.”

He looked at Noah. “You said you replay games in your head. Does that help?”

“Sometimes,” Noah said. “Like if I’m thinking about what I should’ve done instead. But other times I’m just… torturing myself.”

“So there is useful reflection,” the old man said, “and there is useless self-attack. The same wandering ability, used in different ways.

“Let me tell you a little story.”

The Story of the Two Dogs

“In a small village,” he began, “a farmer had two dogs.

“One was a wild, excited dog. It ran all over the fields, chased birds, barked at clouds. It woke up the baby, wasted the farmer’s time, and sometimes knocked over the milk.

“The other dog also loved to run. But the farmer had trained it. It still had energy, but it knew where to go. It chased foxes away from the chickens. It herded the sheep when they wandered near the road. It helped rather than harmed.

“The farmer’s neighbor asked, ‘How did you get one dog to help you and one to cause trouble?’

“The farmer replied, ‘I fed and trained one. I just let the other do whatever it wanted.’

“Our wandering mind is like those dogs. It will run. The question is: have we trained it to be helpful, or do we let it run wild?”

Leah stared at the basketball court below. “So when I lie awake imagining every possible way an exam can go wrong…”

“That,” the old man said, “is the untrained dog. Lots of running. No real help.”

“And when I think back on a game and see what I can actually work on at practice?” Noah asked.

“That is the trained dog,” he nodded. “Same energy. Different direction.”

He folded his hands.

“So we’ve covered:

• Mental strength: focused attention.

• Mental endurance: the wandering mind.
  

“There is one more piece. The one most people forget entirely.”

3. Mindful Presence — Mental Flexibility

The old man grew quiet for a moment. A breeze moved the leaves around them.

“Have you ever had a moment,” he asked, “where everything went oddly quiet inside? Maybe just for a few seconds?”

Leah frowned thoughtfully. “Sometimes when I’m dancing. Like, I stop thinking about the moves and I’m just… there.”

“For me it’s right after a race,” Noah said. “Not during—the middle is just pain and panic. But right after, when I’m lying on the grass, staring at the sky. Brain is empty for a bit.”

He nodded.

“That quiet, that sense of simply being present… that is a glimpse of mental flexibility.

“It’s your ability to:

• Sit or stand without immediately chasing every thought.

• Feel something—nervousness, anger, excitement—without being dragged around by it.

• Notice, ‘Oh, my mind is wandering,’ and gently bring it back.
  

“Most people think of this as peace, calm, or mindfulness. But you can simply call it: being here, on purpose.”

He drew a little circle on the bench with his finger.

“Imagine your thoughts are like the weather: clouds, wind, rain, sun. Constantly changing. Most of us live like we are the weather: ‘I am angry. I am anxious. I am useless.’ We completely blend with whatever passes through.

“This third capacity lets you sit more like the sky: big enough for all of it, not destroyed by every storm.”

Leah watched his finger trace the circle.

“And meditation trains that?” she asked.

“Yes,” he said. “Just as stretching trains flexibility, meditation trains the ability to rest in awareness. To see thoughts and feelings without getting tangled in all of them.”

Why No One Teaches This

Noah leaned back, hands behind his head.

“If it’s so important,” he said, “why don’t we learn this in school? We do PE. We do health class. We never do ‘mind class.’”

The old man smiled sadly.

“Some schools are starting. Slowly. But for a long time, in many places, people assumed that your mind was just… something you had. Like eye color. You used it to study or work, but you didn’t train how it relates to itself.

“There are three main reasons, I think.”

He held up one finger.

“First, you can’t see it. Nobody gets a six-pack of mindfulness. You can’t post a selfie of ‘my brain after ten minutes of meditation.’ So it doesn’t get much attention.”

Second finger.

“Second, it feels like doing nothing. Sitting still, watching your breath, noticing thoughts drift by. Compared to winning a game or solving a hard problem, it doesn’t feel very heroic.”

Third finger.

“Third, progress is subtle. You don’t suddenly float off the ground. Instead, you shout at your sibling a little less. You fall asleep a little faster. You recover from a bad day a little more quickly. These are real changes—but they’re quiet.”

He shrugged.

“And in a culture that loves loud results, quiet progress doesn’t always get much respect.”

Leah nodded slowly. “That actually makes a lot of sense.”

The Three Musicians

“Let me offer another little story,” the old man said.

“Imagine a band with three musicians:

• A drummer: loud, powerful, keeps the beat.

• A guitarist: fast, flashy, plays exciting solos.

• A bassist: quiet, steady, holds the whole song together.
  

“Most people notice the drummer and the guitarist. They get the attention, the applause. But if you remove the bass, the whole song feels thin and unstable, even if you can’t quite say why.

“In your inner world:

• The drummer is your wandering mind, always keeping some rhythm going in the background.

• The guitarist is your focused attention, doing the impressive solos.

• The bass is your mindful presence, the quiet sense of ‘I am here,’ holding everything together.
  

“Most of our training and praise goes to the guitarist and sometimes the drummer. But without the bass, the band never sounds right.”

Noah looked out over the park.

“So meditation is… bass practice,” he said.

The old man laughed. “Yes. Sitting there, practicing ‘boring’ notes that, over time, make the whole song of your life sound different.”

A Tiny Experiment

“Want to try something now?” he asked.

Leah glanced at Noah. “Like, actual meditation?”

“Just one minute,” the old man said. “You don’t even have to close your eyes if that feels weird. Think of it as… a science experiment on your own mind.”

They hesitated, then nodded.

“Good,” he said. “Sit comfortably. Feet on the ground. Hands resting on your legs.

“Now: pick one thing to pay attention to. You can choose:

• The feeling of your breath in your nose or chest.

• The sounds around us.

• Or the way your hands feel resting on your legs.
  

“Got one?”

“Breathing,” said Leah.

“Sounds,” said Noah.

“Perfect. For the next minute, just notice that thing.

“When a thought appears—‘This is dumb,’ ‘What’s for dinner,’ ‘I hope nobody sees me’—don’t fight it. Don’t chase it. Just think: ‘There’s a thought,’ and gently return to your breath or sounds.

“The goal is not to have no thoughts. The goal is to notice when you wander and come back.”

He took out his old watch, set a minute, and nodded.

“Ready? Begin.”

The park went on as usual. A ball thudded against the court down the hill. Somewhere a dog barked. The breeze moved over their faces.

Leah’s eyebrows smoothed a little. Noah’s bouncing legs gradually stilled.

After a minute, the old man said softly, “Okay. Stop.”

Noah blew out a long breath.

“That was… harder than I thought,” he said. “My brain would not shut up.”

“The brain’s job isn’t to shut up,” the old man said. “It’s to think. You wouldn’t get mad at your heart for beating. The point isn’t to silence the mind. It’s to change your relationship to it.”

Leah opened her eyes.

“At first I kept thinking ‘I’m doing this wrong,’” she admitted. “Then I remembered what you said and just went back to the breath. It was… I don’t know. Less panicky than usual.”

The old man smiled.

“And that is the entire principle,” he said. “You wandered, you noticed, you returned. You just completed one full set in the mental gym.”

A Life With and Without Training

“Let me tell you two futures,” he said. “Both completely possible for anyone.”

He held out one hand.

“In the first future, you never train this third capacity. You only push mental strength and mental endurance.

“You get very good at focusing when you have to—exams, deadlines, games. Your wandering mind never stops, constantly replaying the past and guessing the future. Some of that helps you plan. A lot of it keeps you stressed. When strong emotions come, they grab you by the throat. You believe every harsh thought your mind throws at you.

“You might still succeed by many external measures—grades, career, achievements—but inside, you often feel you’re being dragged around by an invisible dog.”

He lifted his other hand.

“In the second future, you practice, just a little.

“Five minutes here, three minutes there, like we just did. You learn to notice thoughts instead of instantly believing them. You catch yourself before you say something cruel in a fight. You still have stress, but it doesn’t own you quite as completely.

“You still study, still train, still dream. But there is a quiet bass line underneath it all: ‘I am here. I can watch this. I can choose my next move.’”

He looked from Leah to Noah.

“Same world. Same phones, same exams, same breakups and disappointments. Very different experience from the inside.”

What Practice Might Look Like

Leah folded her arms, thoughtful.

“How much would we have to do?” she asked. “Like, to make any difference?”

The old man shrugged lightly.

“If you went to the gym for three minutes a day, would you become a bodybuilder?” he asked.

“No,” she said.

“But would you be stronger at the end of the year than if you did nothing at all?” he continued.

“…Probably,” she admitted.

“The mind is similar. Tiny, regular moments can shift things more than you expect.

“You might start with:

• Two minutes before bed, noticing your breath.

• One minute on the bus, just listening to sounds without opening your phone.

• A few deep breaths before an exam, feeling your feet on the floor.
  

“That’s it. No incense required.”

Noah smirked. “No chanting either?”

“Optional,” the old man said, smiling. “You can chant ‘I hope I don’t fail this test’ if you like, as long as you’re noticing that you’re doing it.”

They laughed.

Choosing Your Training

The old man looked at them kindly.

“You don’t have to meditate,” he said. “Just like you don’t have to stretch. You can live with tight muscles and a tight mind. Many people do.

“But you should at least understand the choice.

“If you train only focus and wandering, you might become brilliant and exhausted. If you add this third training—mindful presence—you give yourself a different kind of strength: the ability to stay steady in a noisy world.”

He tapped the bench lightly.

“Just remember this:

• Mental strength is your focus.

• Mental endurance is your wandering thought stream.

• Mental flexibility is your ability to be here and notice it all.
  

“And meditation is simply the practice of sitting down with all three and saying: ‘Let’s learn to work together.’”

Leah stood up, stretching.

“Okay,” she said. “Two minutes a day. I’ll try. If I lose my mind, I’m blaming you.”

“If you find your mind,” the old man said, “you can blame me as well.”

Noah put his headphones back around his neck.

“I’ll try it after practice,” he said. “When I’m lying on the grass. Instead of just roasting myself about every mistake.”

The old man nodded, pleased.

“That,” he said, “is a very good place to start. When the body is tired, the mind often becomes honest.”

They started down the hill together—two teenagers with somewhere to be, and one old man settling back onto his favorite bench.

Behind them, the park continued: balls bouncing, dogs barking, the endless soundtrack of people’s lives.

Inside two young minds, something new had joined the band—a quiet, steady bass line, barely audible at first, but ready to grow stronger with every small, simple practice.

Gemini 3 Pro Thinking

The smell of cedar and old varnish filled the small workshop where Elias spent his days. He was seventy years old, with hands rough as bark and eyes that seemed to hold a quiet, steady light.

Leo and Maya, both seventeen, sat on stools near the woodstove. They looked like typical high school seniors: tired, vibrating with a low-level anxiety, and constantly checking their pockets where their phones buzzed.

“You two look like you’ve been carrying the world,” Elias said, running a piece of sandpaper over the leg of a chair.

“Exam week,” Leo groaned, rubbing his temples. “My brain feels like it’s going to turn into mush.”

“And I can’t sleep,” Maya added, picking at a loose thread on her jeans. “Even when I finish studying, I lay there and my brain just... loops. Over and over.”

Elias blew the dust off the wood. “It sounds to me like you are training recklessly.”

“Training?” Leo asked. “I haven’t been to the gym in weeks. I don’t have time.”

“I’m not talking about your biceps, Leo. I’m talking about the muscle between your ears.” Elias set down the sandpaper. “You know, the mind has a physique, just like the body. And just like the body, it needs three things to be healthy: Strength, Endurance, and Flexibility.”

He walked over to a chalkboard he used for sketching furniture designs and drew a triangle.

The Heavy Lifter (Strength)

“First,” Elias said, pointing to the left corner, “is Strength. Leo, when you go to the gym, why do you lift weights?”

“To get stronger. To be able to move heavy stuff.”

“Exactly,” Elias nodded. “Your mind has strength too. It is called Attention. When you solve a calculus problem, or learn a language, or focus on a complex essay, you are ‘lifting heavy’ with your mind. It is good. It builds a powerful intellect.”

He looked at Leo. “But imagine a weightlifter who lifts for twelve hours straight, never putting the weight down. What happens?”

“He tears a muscle,” Leo said. “Or drops the weight on his foot.”

“Precisely. You are treating your attention like a machine that never overheats. But there is also the matter of junk volume.” Elias leaned in. “When you are too tired to study, what do you do?”

“Watch videos,” Leo shrugged. “Scroll.”

“That,” Elias said, “is like lifting a one-pound weight ten thousand times. It is not heavy enough to make you strong, but it is repetitive enough to wear out your joints. It is low-quality training.”

The Runner in the Smog (Cardio)

Elias turned to Maya. “Now, you said your mind loops. That brings us to the second point: Endurance. Or, as I like to call it, The Traveller.”

He drew a squiggly line on the board.

“When we aren’t focusing hard, our minds wander. This is our ‘default’ state. It is like cardio. It’s a low-intensity jog that can go on all day. It happens when you walk to school, or shower, or eat lunch.”

“That’s when the anxiety hits,” Maya said quietly.

“Imagine a runner,” Elias said. “If a runner jogs through a beautiful forest, they return energized. They have planned their route, enjoyed the view, maybe solved a problem in their head. That is good mental cardio—reflection, planning, dreaming.”

Elias’s expression grew serious. “But what if that runner decides to jog through a tunnel filled with toxic smoke? They are still running, but they are poisoning themselves.”

“That’s the worry,” Maya realized.

“Yes. Rumination. Replaying an embarrassing moment from two years ago. Stressing about a test that hasn’t happened yet. That is running a marathon in smog. You are exhausting yourself, Maya, but you aren’t going anywhere.”

The Oak and The Willow (Flexibility)

Elias picked up a thick strip of dried oak from the floor. “Catch,” he said, tossing it to Leo. “Try to bend that.”

Leo gripped it and strained. The wood barely moved. “I can’t. It’s too hard.”

“It is strong,” Elias corrected. “But it is rigid.”

Then Elias walked to the open window where a willow tree grew outside. He cut a fresh, green branch and handed it to Maya. “Bend that.”

Maya bent the branch into a complete circle. When she let go, it sprang back, undamaged.

“This,” Elias said softly, “is the third element. Flexibility. And this is the one your world has forgotten.”

“In the mind, flexibility is the ability to simply be,” Elias continued. “To sit in the present moment without judging it, without fighting it, and without trying to fix it. We call this meditation.”

Leo scoffed slightly. “Sitting still? That feels like doing nothing. How is that training?”

“Let me tell you a story,” Elias said, sitting back on his stool.

“Once, a great storm swept through the forest. The winds howled like demons. The mighty Oak tree puffing out its chest said, ‘I am strong! I have deep roots! I will not move!’

The Willow tree looked at the wind and said, ‘I cannot stop you, so I will move with you.’

The wind blew harder. The Oak fought the wind with all its strength (Attention). It braced itself. But the wind was too chaotic. With a sound like a cannon shot, the Oak snapped in half.

The Willow, however, yielded. It bowed low, letting the wind pass over it. It didn’t try to analyze the wind (Strength), and it didn’t run from the wind (Cardio). It simply felt the wind and let it pass.

When the storm ended, the Oak was destroyed. The Willow stood back up, shook its leaves, and drank the rain.”

Elias looked at them. “You two are training to be Oaks. You are building massive mental strength in school. You are running endless mental marathons with your worry. But you have no flexibility. So when the storm comes—a breakup, a failure, a tragedy—you are at risk of snapping.”

“So... meditation is just stretching?” Leo asked, looking at the rigid piece of oak in his hands.

“It is the practice of putting down the weight,” Elias said. “It is stopping the runner. It is the only time in your day when you are not ‘doing’ or ‘going.’ You are just sitting. It feels difficult because you are stiff. Touching your toes is hard when you have never stretched before.”

The Lesson

The workshop was quiet. The frantic energy of the teenagers had settled into something calmer.

“The world you are growing up in wants you to be a machine,” Elias said gently. “It wants output. It wants grades. It wants likes. It tells you that sitting still is lazy.”

He stood up and dusted off his apron.

“But a machine has no flexibility. A machine grinds its gears until it breaks. To be a human—a healthy human—you must learn the art of the Willow. You must learn to sit for ten minutes, watch the storm of your thoughts, and let them pass without fighting them.”

“How do we start?” Maya asked.

“Start small,” Elias smiled. “Five minutes a day. No phone. No music. Just sit. When your mind tries to pick up a weight (thinking), put it down. When your mind tries to start running (worrying), tell it to stop. Just sit. It will be the hardest training you do, but it will save you when the storm comes.”

Leo looked at the oak wood, then at Maya. “Maybe we can try it? Before we go back to studying?”

Maya nodded. “I think I’d like that.”

And there, amidst the smell of sawdust and the ticking of the workshop clock, two young oaks began to learn the way of the willow.

Grok 4.1 Thinking Beta

The sun had already slipped behind the western ridge, but the sky was still the color of a fresh bruise.

Aria and Jax sat on a log by the dying campfire, shoulders touching because the night was turning cold.

They had come up the mountain to “get away from everything,” but everything had followed them in their pockets and inside their skulls.

An old man stepped out of the dark as if the pines themselves had decided to speak.

His beard was white as birch bark, his eyes bright as a hawk’s.

He carried no pack, only a walking stick worn smooth by decades of mountains.

“Mind if I borrow some of your fire?” he asked.

His voice was low, warm, the kind you feel in your ribs.

Jax shrugged. “It’s almost out anyway.”

The old man—Elias, he told them—sat cross-legged on the ground without waiting for more invitation.

He fed the fire three small sticks and it leapt back to life like it recognized him.

“You two of you,” he said, “are strong in some places and brittle in others.

I can see it in the way you keep checking those rectangles of light in your hands.”

Aria laughed, a little embarrassed, and shoved her phone into her jacket.

“We’re fine,” Jax muttered. “Just stressed.”

Elias smiled. “Stressed is what a bowstring says right before it snaps.”

He leaned forward, firelight flickering across his face.

“Let me tell you about three young warriors I once knew.

The first was called Iron Arm.

He could bend iron bars with his hands and lift boulders that took three ordinary men ropes and levers.

Villages cheered when he walked past.

But one winter he had to cross a narrow footbridge in a storm.

The wind pushed; he could not bend.

He fell into the gorge and was never seen again.

The second was called Wind Foot.

He could run from dawn to dusk without tiring, could outrun horses on the open plain.

Everyone said he would live forever.

But when raiders came to his village he could not lift a sword heavy enough to matter.

He ran, yes.

He lived.

But his people did not.

The third was called Willow.

She was not remarkably strong, and she tired like anyone else after a long march.

But she could twist and flow like a dancer in battle.

Arrows missed her.

Swords cut only air.

When the storm came she bent with the bridge and crossed safely.

When raiders came she slipped between them like smoke.

She lived a very long life, and people still tell her stories.

Strength without flexibility is pride.

Speed without strength is fear.

Only the one who trained all three parts became unbreakable.”

Elias poked the fire. Sparks rose like tiny stars.

“Your mind is the same,” he continued. “Exactly the same.

Most young people today train only two parts and wonder why they feel half-dead inside.

The first part is the blacksmith’s forge—your focused mind.

This is the power to hammer one thing until it takes shape: studying for exams, practicing free throws, writing a song, learning code, fixing an engine.

When you give something your full attention, you are swinging the hammer.

Good strikes make the blade sharp.

But if all you ever do is swing the hammer—grind, hustle, achieve, scroll, game, hustle some more—your arm gets huge and your back seizes up.

I’ve seen seventeen-year-olds with the soul of fifty-year-old machines, rusted tight.

The second part is the wandering horse.

When you walk home from school, wash dishes, lie in bed—your mind runs free.

This is not laziness; this is the long, slow miles that build endurance in your spirit.

A good wander reviews the day and learns from it.

A good wander dreams about the person you might become and feels the excitement.

A good wander invents stories in the shower that later turn into real things.

But most wandering horses today are chased in circles by wolves of worry, regret, comparison, rage.

They run and run and run until they drop dead from exhaustion, and the rider never gets anywhere.

And then there is the third part—the still pond.

Almost nobody trains this one.

Imagine a lake high in the mountains.

When it is stormy, the water is muddy; you cannot see your own face.

When the wind dies and the surface becomes glass, you see not only your reflection but the sky behind you, the eagle overhead, every pebble on the bottom.

That stillness is not emptiness.

It is completeness.

We call the training ‘meditation,’ but that word sounds like homework to you.

So call it something else.

Call it stretching for the soul.

Call it giving your mind a bath.

Call it the place where you remember you are alive and not a machine.

Five minutes a day is enough to start.

Sit.

Feel your breath go in and out like waves on a shore.

Thoughts will come—let them sail past like clouds.

You do not have to fight them.

Fighting thoughts is just another way thoughts win.

Do this and something strange happens.

The blacksmith’s arm stops aching.

The wandering horse finds better pastures.

The storms still come, but you are no longer the storm.

I knew a girl your age once,” Elias said softly, looking at Aria. “She was brilliant—top of every class, captain of every team.

But inside she was screaming all the time.

One day she sat by a river for seven minutes and simply listened to the water.

She cried for no reason and every reason.

After that, the screaming quieted.

She still worked hard, but now the work belonged to her instead of owning her.

And I knew a boy your age,” he said, turning to Jax, “who could not sleep because tomorrow was always attacking him.

He started watching his breath every night before bed.

Not to fall asleep—just to watch.

Sleep came as a friend instead of an enemy.

He began to dream again.

Real dreams.

The kind that change lives.”

The fire had settled into slow red coals.

Elias stood.

His knees did not creak.

“Tonight,” he said, “when you zip into your sleeping bags, do not open the glowing rectangles.

Lie on your backs.

Feel the weight of your body on the earth.

Notice one breath.

Then another.

When the mind runs, smile at it the way you’d smile at a puppy that escaped the yard.

Then come back to the breath.

Do it for seven nights.

If on the eighth morning you do not feel even a little lighter, come find me and I will give you my walking stick as payment for your trouble.”

He turned to leave, then paused.

“One last thing.

The world will never give you permission to be still.

It makes no money from your peace.

So you must steal the minutes like a thief.

But once you taste it—real quiet—you will steal more, gladly.

Because a river is only free when it remembers it is also the ocean.

Good night, young willow trees.

Grow bent, grow strong, grow deep.”

He walked into the dark.

The pines closed behind him like a curtain.

Aria and Jax sat without speaking for a long time.

Then, without planning it, they lay back on their sleeping bags and stared up at the stars.

For the first time in months, the silence did not feel like something they had to fill.

Somewhere inside their chests, very gently, something stretched.

Opus 4.1

The Garden of Three Paths

The old clockmaker’s shop sat wedged between a bustling café and a vintage bookstore, easy to miss if you weren’t looking for it. Maya and James had walked past it dozens of times before Mr. Chen had invited them in one rainy afternoon, offering shelter and tea while they waited for the storm to pass.

That was three months ago. Now they came every Tuesday after school, drawn by something they couldn’t quite name—perhaps the way the old man’s stories seemed to answer questions they hadn’t known how to ask.

Today, Maya slumped in her usual chair, dark circles under her eyes. “I got into Berkeley,” she announced without joy.

James looked up from the watch mechanism he’d been studying. “That’s incredible! Why do you look like someone died?”

“Because I can’t sleep. I can’t stop thinking about everything—will I be good enough, what if I chose the wrong major, what if—” She rubbed her temples. “My mind won’t shut off.”

Mr. Chen set down the delicate gears he’d been cleaning. His weathered hands, steady despite his eighty years, poured tea for all three of them. “Ah,” he said simply. “Maya, you’ve been training for a marathon using only your arms.”

James snorted. “What does that even mean?”

The old man smiled. “Tell me, what makes an athlete truly healthy?”

“Being strong?” James ventured.

“Being able to run far?” Maya added.

“And?”

They looked at each other, then Maya said slowly, “Being flexible?”

“Precisely. Three pillars, each essential. Too much focus on one creates imbalance.” Mr. Chen moved to his workbench, where an intricate clock face lay disassembled. “Your minds are no different. But first, let me tell you about the village of runners.”

The Village of Runners

Mr. Chen settled into his story-telling voice, the one that made the shop seem to fade away.

“There was once a village obsessed with running. From childhood, everyone ran. They had the most magnificent legs you’d ever seen—powerful, tireless. They could run from sunrise to sunset without stopping. They were proud of this, believing themselves the healthiest people in all the land.

“One day, a tree fell across their main path. It was too heavy to lift, too low to run under comfortably. The villagers had to crawl beneath it or climb over it, but their backs ached terribly—they’d never learned to bend. Their arms, weak from neglect, could barely pull them over. Some injured themselves trying; others took long detours to avoid the tree entirely.

“A traveling merchant, seeing their struggle, showed them how to stretch, how to build strength in their arms and backs. But the villagers resisted. ‘We are runners,’ they said. ‘This other training would take time from our running.’

“The tree remains there to this day.”

Maya frowned. “So they just... kept being limited?”

“They chose their limitation,” Mr. Chen corrected gently. “Now, your mind has three aspects, much like the body. James, you’re quite good at the first—focused attention. I’ve watched you work on these mechanisms for hours without wavering.”

James straightened slightly, pleased.

“But when you leave here, what happens?”

“I... what do you mean?”

“Do you ever just sit? Just observe? Or does your mind immediately seek the next task, the next problem to solve?”

James shifted uncomfortably. “I like being productive.”

“And Maya,” Mr. Chen continued, “your second aspect—your wandering mind—is incredibly active. You think, you plan, you reflect. But like a river that never stops flowing, it’s wearing away at your banks.”

The Three Gardens

Mr. Chen led them to the back room they’d never entered before. To their surprise, it opened onto a small courtyard with three distinct garden areas.

The first was filled with puzzle boxes, mechanical devices, and books. “The Garden of Focus,” he explained. “Where the mind learns to direct its strength.”

The second held comfortable chairs facing windows, wind chimes, and journals. “The Garden of Wandering, where thoughts flow and connect.”

The third was simplest—cushions on smooth stone, a small fountain, nothing else. “The Garden of Presence.”

“This one’s empty,” James observed.

Mr. Chen chuckled. “Is it? Sit there for five minutes and tell me what you notice.”

James sat, fidgeting almost immediately. After barely two minutes, he stood. “I can’t just sit there doing nothing.”

“Exactly your problem,” Mr. Chen said kindly. “You’ve never developed your mental flexibility. Maya, you try.”

Maya lasted longer but soon her face showed distress. “My thoughts are so loud here. Without distraction, they’re overwhelming.”

“Now you understand,” Mr. Chen said, “why I invited you both here. You’re each strong in different ways, but incomplete. Like a stool with one tall leg and two short ones—it will always wobble.”

The Merchant’s Three Sons

They returned to the main shop, and Mr. Chen began another story while his hands worked on a pocket watch.

“A wealthy merchant had three sons. When deciding who would inherit his business, he gave each son the same amount of money and told them to develop themselves for one year.

“The first son hired the best tutors, studying mathematics, languages, and strategy from dawn to midnight. He learned five languages and could calculate complex equations in his head. But the constant study exhausted him. He developed headaches, couldn’t enjoy simple pleasures, and despite his knowledge, made poor decisions when tired—which was always.

“The second son traveled the world, experiencing everything, letting his mind wander from adventure to adventure. He gathered wonderful stories and creative ideas but never developed the discipline to implement them. His journals overflowed with half-formed dreams that never became reality.

“The third son puzzled everyone. He spent his mornings in focused study, his afternoons walking and reflecting, and his evenings in meditation. People mocked him—’You’re wasting a third of your time sitting still!’

“After a year, the merchant tested all three. He presented them with a complex business crisis requiring immediate decision. The first son knew all the theory but was too mentally exhausted to apply it well. The second had creative solutions but couldn’t focus long enough to see them through. The third listened calmly, focused completely on the problem, drew from his reflections to find creative solutions, and had the mental clarity to know which path was wisest.

“’How?’ his brothers asked.

“’You strengthened only one aspect of your minds,’ he replied. ‘I developed all three. My meditation didn’t make me weaker—it made everything else stronger.’”

The Practice

Maya leaned forward. “So meditation is like... mental flexibility training?”

“Tell me,” Mr. Chen said, “what happens when you stretch regularly?”

“You become more flexible?” James offered.

“Yes, but what else? What happens to your strength training, your running?”

Maya’s eyes widened. “They improve. You have better form, fewer injuries, better recovery.”

“Precisely. Meditation—true mindfulness—isn’t about emptying your mind or achieving bliss. It’s about developing the capacity to observe without being pulled into the current. Watch your thoughts like you’d watch clouds passing. They’re there, they’re real, but you don’t have to chase each one.”

“But how?” Maya asked, frustration creeping in. “Every time I try, my mind goes everywhere.”

Mr. Chen pulled out an old snow globe, shaking it vigorously. The snow swirled chaotically. “This is your mind now, Maya. Always shaking, always swirling.” He set it down. Gradually, the snow settled, the water cleared. “Meditation isn’t about stopping the snow. It’s about letting it settle. But every time you shake the globe checking if it’s clear yet...”

“You stir it up again,” she finished.

The Mountain Climber’s Lesson

“But sitting still feels like wasting time,” James protested. “There’s so much to do, to learn.”

Mr. Chen nodded. “I once knew a mountain climber, incredibly ambitious. He would race up mountains, proud of his speed. One day, he attempted the highest peak in the range. Halfway up, altitude sickness struck. He had to descend, defeated.

“He met an old sherpa who told him, ‘You climb like you’re fighting the mountain. The mountain doesn’t care about your schedule.’

“The sherpa taught him to climb differently—periods of effort followed by complete rest, not just physical but mental. Sitting, breathing, being present with the mountain rather than conquering it.

“The climber resisted—it felt too slow. But he tried it. To his amazement, he reached summits he’d never achieved before. The rest wasn’t weakness; it was part of the strength. His body adapted better, his mind stayed clear at altitude, and he actually enjoyed the journey instead of just enduring it for the summit.”

James was quiet, processing.

“Your generation,” Mr. Chen continued, “has been taught to optimize everything, to be productive every moment. But a bow kept always strung will lose its power. A field never left fallow will deplete. And a mind never given rest will eventually rebel—through anxiety, through insomnia, through burnout.”

The Three Practices

“So what do we actually do?” Maya asked.

Mr. Chen handed them each a small notebook. “Three practices, one for each garden. Choose based on what you need most, but don’t neglect any entirely.

“For Focus: Choose something requiring complete attention. Mathematics, music, coding, detailed art. But—and this is crucial—do it without distraction. No phone, no music unless that IS the practice. One thing, fully engaged. Start with twenty minutes if an hour seems impossible.

“For Wandering: Schedule time for quality reflection. Walk without podcasts. Sit with a journal. Let your mind process, but guide it gently toward construction rather than destruction. If you notice yourself spiraling into worry, ask: ‘Is this thought helping me prepare, or just making me suffer?’

“For Presence: Start with just five minutes daily. Sit, breathe, notice. When thoughts arise—and they will—don’t fight them. Notice them like you’d notice a sound outside, then return to breath. It’s not about perfect stillness; it’s about practicing return.”

The Story of the Anxious Archer

Seeing Maya’s skeptical expression, Mr. Chen shared one more tale.

“There was an archer, talented but anxious. In practice, she never missed. In competition, her mind raced—what if she missed, what would people think, what about the archer next to her who looked so confident? Her arrows flew wild.

“A master archer watched her and said, ‘Your problem isn’t your aim; it’s that you’re shooting three arrows at once—one at the target, one at your fear, and one at the future.’

“He taught her to meditate. Not to empty her mind, but to notice when she was shooting multiple arrows. Through practice, she learned to recognize the moment her mind split, to breathe, to return to just one arrow, one target, one moment.

“She didn’t become fearless. The thoughts still arose. But she learned to let them pass through without gripping them. Her accuracy in competition finally matched her practice. More importantly, she began to enjoy the sport again.”

Maya’s eyes glistened slightly. “That’s me with tests. I know the material, but my mind...”

“Races ahead, imagining failures that haven’t happened, creating suffering from thin air,” Mr. Chen finished gently.

The Modern Challenge

“Your generation faces something unique,” Mr. Chen said, cleaning his tools methodically. “Never before have humans had such constant stimulation available. Your phones—miraculous devices—also mean you never have to sit with discomfort, with boredom, with your own thoughts.

“But boredom is where creativity is born. Discomfort is where growth happens. And sitting with your thoughts, learning to observe them without being controlled by them—this is where wisdom begins.”

James pulled out his phone reflexively, then caught himself and put it away, embarrassed.

“Don’t be ashamed,” Mr. Chen smiled. “It’s designed to capture your attention. But ask yourself: are you using it, or is it using you? Can you sit through a meal without checking it? Can you wake up without immediately reaching for it? These aren’t moral failings—they’re signs of an untrained aspect of mind.”

The Water and the Cup

Mr. Chen poured more tea, deliberately, mindfully. “Watch,” he said. “The water takes the shape of the cup, but it remains water. Your thoughts, your wandering mind, your focused attention—they’re all water. Meditation teaches you to be the cup as well as the water. To hold your thoughts without becoming them.”

“That doesn’t make sense,” James said.

“Not to the mind that only knows doing,” Mr. Chen agreed. “But sense isn’t always found through thinking. Sometimes it’s found through practice, through experience. You learned to ride a bicycle not by understanding the physics, but by getting on and falling until suddenly, you weren’t falling anymore.”

The Assignment

As the afternoon light faded, Mr. Chen gave them their task.

“For one week, practice all three aspects daily. James, you’ll meditate for ten minutes each morning—and I mean truly sit, not plan your day while sitting. Also, take one walk daily without any goal, let your mind wander constructively.

“Maya, you’ll do one focused task daily without any multitasking—perhaps mathematics or art. And yes, you’ll also meditate, but differently. Your practice is to notice when your mind spirals and gently return to present, even during daily activities.”

“What if we’re bad at it?” Maya asked.

“Then you’re doing it correctly,” Mr. Chen laughed. “Everyone is ‘bad’ at meditation when they start. That’s like saying you’re bad at lifting weights because they feel heavy. The weight is supposed to feel heavy. Your mind is supposed to wander. The practice isn’t in preventing that—it’s in noticing and returning.”

The Clock’s Secret

Before they left, Mr. Chen showed them the clock he’d been repairing. “See these three mechanisms? The mainspring provides power—like your focused attention. The escapement regulates the energy—like your wandering mind, processing and distributing. And the balance wheel maintains rhythm—like mindfulness, keeping everything in harmony.

“Remove any one, and the clock stops. Neglect maintenance on any part, and time distorts. But when all three work together, properly maintained...” He wound the clock. It ticked with perfect rhythm.

The Return

A week later, they returned. James entered more calmly than usual, moving deliberately rather than his typical rush.

“The meditation is hard,” he admitted. “But I noticed something. After sitting still, even for just ten minutes, my focus during work is... cleaner? Like someone cleaned my mental glasses.”

Maya nodded eagerly. “I know what you mean! I still have anxious thoughts about college, but now I sometimes catch myself spiraling and can ask, ‘Is this helpful or just noise?’ Not always, but sometimes. And that’s new.”

Mr. Chen smiled. “You’ve discovered something important. Mental fitness isn’t about perfection or constant peace. It’s about developing the capacity to work with your mind rather than being worked by it.”

The Final Story

“Let me leave you with one last story,” Mr. Chen said. “A student asked a wise teacher, ‘Master, you seem so peaceful. Have you conquered your mind?’

“The teacher laughed. ‘Conquered? No. My mind still produces thoughts like clouds produce rain. The difference is I’ve learned to build good shelters and when to go walking in the rain. I’ve learned when thoughts are weather to wait out and when they’re climate requiring action. But conquered? The mind isn’t an enemy to defeat—it’s a powerful horse to train. And training never really ends; it just becomes more refined.’

“’Then what’s the point?’ the student asked.

“’The point,’ the teacher said, ‘is that an untrained horse goes where it wants, when it wants, dragging you along. A trained horse takes you where you need to go. Both are still horses. One is a partnership; the other is chaos.’”

The Continuation

As Maya and James prepared to leave, Mr. Chen handed them each a small river stone, smooth and plain.

“Keep these,” he said. “When you feel your mind racing, hold it. Feel its weight, its texture, its temperature. Let it anchor you to the present. Not as magic, but as reminder—you always have access to this moment, this breath, this choice of where to direct your attention.”

“Will it get easier?” Maya asked.

“Different,” Mr. Chen corrected. “Your capacity will grow. You’ll notice patterns sooner, return to balance faster. But life will also bring new challenges requiring deeper practice. This isn’t a problem to solve once—it’s a relationship to tend daily.”

James pocketed his stone thoughtfully. “It’s like maintaining these clocks, isn’t it? Daily attention, small adjustments, regular care.”

“Now you’re beginning to understand,” Mr. Chen said warmly. “Your minds are the most sophisticated instruments you’ll ever own. They deserve at least as much care as we give these simple machines.”

As they left, Maya turned back. “Mr. Chen, how long did it take you to develop all three aspects?”

The old man’s eyes crinkled with humor. “I’ll let you know when I’m finished.”

Walking away, James and Maya felt something had shifted. The world looked the same, but they had new eyes for seeing it—and more importantly, new ways of seeing their own seeing. The journey toward mental fitness had begun, not with grand gestures, but with simple awareness: that the mind, like the body, needs strength, endurance, and flexibility. And that developing all three was not just possible, but essential.

The clock tower chimed five o’clock. For the first time, they both stopped to listen to it—really listen—until the last echo faded into silence.

Opus 4.5

The Three Gardens

The morning sun had just crested the eastern hills when Maya and her cousin Daniel arrived at the old wooden gate. Beyond it, a narrow path wound through wild grasses toward a small house that seemed to grow out of the hillside itself. Wisteria climbed its walls, and somewhere behind it, they could hear the soft music of running water.

“Are you sure about this?” Daniel asked, adjusting his backpack. “We don’t even know this guy.”

“Gran said he helped her through the hardest year of her life,” Maya replied. “She said if we ever felt lost, we should find him.”

Daniel didn’t argue. They had both, in their own ways, been struggling. Maya had spent the past year in a fog of anxiety that no amount of advice seemed to lift. Daniel had dropped out of his first semester of college, unable to explain why except to say that his mind felt like a browser with a hundred tabs open and no way to close any of them. Their grandmother had listened to them both at the family gathering last month, nodded slowly, and written an address on a piece of paper.

The gate swung open silently, and they walked the winding path until they reached a small courtyard where an old man sat on a wooden bench, eyes closed, perfectly still. He looked ancient—his face carved with deep lines, his hair white and thin—but when he opened his eyes at the sound of their footsteps, those eyes were sharp and clear and full of something that might have been amusement.

“Margaret’s grandchildren,” he said. It was not a question. “Sit. I’ve just made tea.”

They sat in his small kitchen, cups warming their hands, while the old man studied them with patient curiosity. His name, he told them, was Edward, though their grandmother had always called him Ned.

“She said you might be able to help us,” Maya ventured. “She said you taught her something important once.”

“Did she tell you what?”

“No. She said we had to find out for ourselves.”

Edward smiled. “Your grandmother was always wiser than she pretended to be.” He rose slowly and gestured toward the door. “Come. Let me show you something.”

He led them around the house to where the sound of water grew louder. There, hidden from the front path, lay three distinct gardens separated by low stone walls.

The first garden was a workshop as much as a growing space. Raised beds held vegetables in neat rows. Fruit trees were carefully pruned and trained along wires. A small toolshed stood open, revealing implements hung in precise order on the walls.

The second garden was wilder. A stream ran through it, feeding a pond where lily pads floated. Wildflowers grew in scattered clusters, and winding paths disappeared around corners into hidden alcoves. A wooden bench sat near the water’s edge.

The third garden was nearly empty. A single ancient tree stood at its center, its broad canopy creating a circle of shade on the grass beneath. There was nothing else—no flowers, no paths, no structures. Just the tree, the grass, and the sky above.

“Three gardens,” Edward said. “I have spent most of my life learning what they have to teach.”

They returned to the bench in the courtyard, and Edward poured more tea.

“Let me tell you about a woodcarver I once knew,” he began. “His name was Thomas, and he could make the most extraordinary things—birds that seemed ready to take flight, faces that held entire lifetimes in their expressions. People came from great distances to commission his work.

“One day, a young apprentice asked him his secret. Thomas led the boy to his workshop and pointed to the tools on the wall. ‘Do you see these chisels? Each one is sharp enough to cut through oak with a single stroke. But sharpness alone means nothing. Watch.’

“He placed a block of wood on his workbench and picked up a chisel. Then, instead of carving, he turned to the apprentice and began chatting about the weather, about the news from town, about a dozen trivial things. All the while, his hands moved carelessly over the wood. Within minutes, he had ruined the block—gouges and scratches everywhere, the beginnings of no recognizable form.

“’You see?’ Thomas said. ‘The tool is just as sharp as before. But a sharp tool without focused attention creates nothing of value. It may even cause harm.’ He picked up a fresh block of wood. This time, he fell silent. His eyes narrowed. His breathing slowed. For the next hour, he did not speak or look up. When he finally set down his chisel, a delicate flower had emerged from the wood, every petal perfect.

“’Attention,’ Thomas told the boy, ‘is the hand that guides the tool. The tool does not care what it cuts. Only attention can create something meaningful from raw material.’”

Edward paused and looked at Daniel. “You told your grandmother your mind feels like too many things happening at once. Too many thoughts, too many distractions, too many voices competing for your attention.”

Daniel nodded slowly.

“The modern world offers endless tools,” Edward continued. “Endless information, endless entertainment, endless stimulation. But it does not teach you how to hold a chisel. It does not teach you how to guide your attention with intention. And so people carve away at their hours, their days, their years—and at the end, they look down and find they have created nothing they meant to create. Just gouges and scratches. The raw material of their life, wasted.”

“So the first garden,” Maya said, glancing toward the vegetable beds and the orderly toolshed, “that’s about focused attention?”

“Very good. That garden requires work. Deliberate effort. Tasks that demand something from you. When I weed those beds, I am not merely growing vegetables—I am training my capacity to direct my mind where I choose. When I sharpen those tools and hang them in their places, I am practicing the art of doing one thing at a time, and doing it completely.”

He leaned forward. “Here is what most people do not understand: attention is not simply something you have. It is something you train. Every time you engage fully with a challenging task—a real conversation, a difficult problem, work that requires your whole mind—you are strengthening your capacity to focus. And every time you give your attention away to something that asks nothing of you, you weaken it.”

“But everyone needs to rest,” Daniel said. “You can’t focus all the time.”

Edward’s eyes sparkled. “Ah. Now you are ready for the second garden.”

They walked to the wilder space, where the stream murmured over smooth stones and butterflies drifted among the wildflowers. Edward led them to the bench by the pond.

“Let me tell you about a river,” he said.

“There was once a village built along a great river. The river provided everything—water for crops, fish for food, a path for trade. The villagers loved the river and depended on it completely.

“But the river had two faces. When the spring rains were gentle, the river flowed calm and clear. It nourished the fields, turned the mill wheels at a steady pace, and carried boats safely to neighboring towns. The villagers would walk along its banks in the evening, watching the water catch the last light of day, and they felt at peace.

“But some years, the rains came too hard and too fast. The river would swell and churn, breaking its banks, flooding the fields, sweeping away anything not anchored down. The same water that gave life could also destroy it. The villagers learned to watch the sky with wary eyes.

“One year, a young woman in the village grew tired of this uncertainty. ‘Why should we live at the mercy of the river?’ she asked. ‘Let us build a great dam and control the water completely. We will release only what we need, when we need it. The floods will end forever.’

“Some villagers thought this was wise. But an old fisherman shook his head. ‘You cannot dam a living river without killing it,’ he said. ‘The floods are destructive, yes. But they also deposit rich soil on the fields. They clear away dead wood and make room for new growth. They remind us that we are part of something larger than ourselves. If you try to control the river completely, you will lose everything it gives you.’

“The young woman dismissed him as old-fashioned. She organized the village, and they built an enormous dam. For several years, everything seemed perfect. The water flowed exactly as they directed. There were no floods, no uncertainty.

“But slowly, things began to change. The fields grew less fertile without the river’s deposits. The fish disappeared, unable to complete their ancient migrations. The water behind the dam grew stagnant and foul. And the villagers, who had once walked the riverbanks in wonder, now saw only a wall of stone and a trickle of water they had tamed into submission.

“In the end, they had to tear the dam down. The river returned to its natural rhythms—sometimes gentle, sometimes wild. The villagers learned to work with it rather than against it. They built their homes on higher ground and their boats more sturdy. They planted crops that could survive occasional flooding. They stopped trying to control the water and started trying to understand it.”

Edward gestured at the stream flowing through the garden. “Your mind has a river too. When you are not focusing on something specific, your thoughts flow on their own—sometimes gently, sometimes in floods. Memories of the past, worries about the future, fantasies, regrets, plans, fears. This stream is always moving. You cannot dam it, and you should not try.”

“So we just let our thoughts run wild?” Maya asked.

“You let them run,” Edward said. “But you also learn to understand them. Look—” He pointed to the pond where the stream widened and slowed. “When a river enters a wide, calm pool, what happens to the sediment it carries?”

“It settles,” Daniel said. “The water clears.”

“Yes. The thoughts your mind carries are like that sediment. When you give yourself space to reflect—real reflection, not distraction—the water clears and you can see what you are carrying. You can examine it. You can decide what to keep and what to release.”

He picked up a fallen leaf and dropped it into the stream. They watched it swirl in a small eddy, then continue downstream.

“There is healthy reflection,” Edward continued. “Thinking about something difficult that happened and asking what you can learn from it. Considering a challenge ahead and preparing yourself to meet it. Allowing your mind to wander into creative possibility, imagining what your life might become. These are the gentle flows that nourish the fields.

“And there is unhealthy reflection. Replaying the same painful moment over and over, not to learn from it but to suffer it again. Worrying about the future endlessly without ever taking action. Telling yourself the same stories of inadequacy until you believe they are the only truth. These are the floods that destroy.”

Maya was quiet for a long moment. “How do you tell the difference? When you’re in it, I mean. How do you know if you’re reflecting or just... drowning?”

Edward smiled gently. “That is exactly the right question. And it brings us to the third garden.”

The third garden was the strangest of all. Just the great tree, the circle of grass beneath it, and an overwhelming sense of stillness. Even the breeze seemed to pause at the low stone wall.

“Sit,” Edward said, lowering himself to the grass beneath the tree. Maya and Daniel joined him, and for a long moment, no one spoke.

“Let me tell you one more story,” Edward said at last. “A story about a lake.

“High in the mountains, there was once a lake so still and clear that travelers came from distant lands just to look at it. On calm days, the lake reflected the surrounding peaks so perfectly that you could not tell where the water ended and the sky began. People said that to look into this lake was to see truth itself.

“A young scholar heard of this lake and made the long journey to find it. When he arrived, he was disappointed. The wind was blowing, and the surface of the water was choppy and broken. He could see nothing but fractured images, shattered reflections.

“’This is no mirror of truth,’ he complained to an old woman who lived nearby. ‘It’s just a lake. The stories were lies.’

“The old woman laughed. ‘The lake is exactly what the stories say. But you cannot see the reflection while the wind blows. You must wait.’

“’Wait for what?’

“’For stillness.’

“The scholar was impatient, but he had come too far to leave without seeing what he had come for. So he waited. For three days, the wind blew and the water churned. The scholar grew frustrated, bored, restless. He paced the shore and cursed his wasted time.

“On the fourth morning, he woke to absolute silence. He walked to the water’s edge and stopped. The lake was perfect glass. The mountains hung suspended in it, every detail precise. And there—in the center of the reflection—was his own face, clearer than any mirror had ever shown him.

“For a long time, he simply looked. He saw things in that face he had never allowed himself to see. Fears he had been running from. Hopes he had been afraid to name. The truth of who he was, beneath all the stories he told himself.

“When the wind returned that afternoon, the scholar began his journey home. He was not a different person than when he had arrived. But he knew himself now. And that changed everything.”

Edward opened his eyes and looked at his young visitors.

“Your mind is like that lake. The wind is always blowing—thoughts, worries, plans, memories, desires. The surface is always moving. And as long as it moves, you cannot see clearly. You cannot tell the difference between healthy reflection and destructive rumination because you cannot see yourself at all. You are too caught up in the motion of the water.”

“So meditation,” Maya said slowly, “is waiting for the wind to stop?”

“Meditation is discovering that you are not the wind. You are the lake.” Edward placed his palm flat on the grass. “When you sit quietly and simply observe your thoughts—not following them, not fighting them, just watching them pass like clouds across the sky—something remarkable happens. You begin to realize that you are not your thoughts. You are the awareness in which thoughts arise.

“This is very hard to understand until you experience it. But once you do, everything changes. When a worried thought arises, you do not have to become worry. You can simply notice: there is worry. When an angry thought arises, you do not have to become anger. You can notice: there is anger. The thought passes, and you remain.”

Daniel frowned. “But I’ve tried meditation. I can’t make my thoughts stop. My mind just keeps going.”

“Of course it does. The goal is not to stop your thoughts. The goal is to change your relationship to them. Every time you notice that your mind has wandered, you are waking up. That moment of noticing—that is the practice. That is the repetition that builds the skill. It does not feel like progress because you expect progress to feel like success. But each moment of awareness, even after a long stretch of distraction, is a small victory.”

He reached up and touched the trunk of the great tree. “Why do you think this garden has only a tree?”

Maya considered. “Because you don’t need anything else. There’s nothing to do here except... be here.”

“Yes. The first garden asks something of you—effort, attention, work. The second garden offers something to you—a place to wander, to reflect, to watch your thoughts flow by. But the third garden asks nothing and offers nothing except presence. And that is exactly what makes it so difficult. We are not trained to simply be present. We are trained to do, to achieve, to improve, to succeed. Sitting still feels like failure because we have no framework for understanding what sitting still accomplishes.”

“So what does it accomplish?” Daniel asked.

Edward was quiet for a moment. “Let me answer with another story.

“There was once a young man who was an excellent archer. He could hit any target, no matter how small or how distant. He won every competition and earned great fame.

“One day, he heard of a master archer who lived in the mountains—a woman so skilled that she had not touched a bow in twenty years, yet was said to be the greatest archer alive. The young man was skeptical, but curious. He made the journey to find her.

“When he arrived, he challenged her to a contest. She agreed. They would each shoot at a leaf hanging from a tree across a deep gorge.

“The young man went first. He drew his bow, took careful aim, and released. His arrow flew true and split the leaf perfectly in half. He turned to the old woman with a proud smile.

“She did not pick up a bow. Instead, she walked to the edge of the gorge—a sheer drop of hundreds of feet to the rocks below. She stepped out onto a narrow branch that extended over the void, walked to its very tip where it bent and swayed beneath her weight, and stood there, perfectly calm, gazing at the distant leaf.

“’Join me,’ she said.

“The young man tried. He stepped onto the branch, felt it bounce and sway, looked down at the dizzying drop, and immediately scrambled back to solid ground, his whole body trembling.

“’Your skill with the bow is impressive,’ the old woman said, walking calmly back to shore. ‘But your skill with yourself is lacking. What good is a steady hand if your heart is in chaos? What good is perfect aim if your mind can be shattered by a change in circumstances? You have trained your body and your technique. But you have not trained your being.’

“The young man understood. He stayed in the mountains for many years, learning not how to shoot but how to be still. When he finally returned to the world, he was no longer interested in competitions. But those who saw him move through life said there was something different about him—a presence, a groundedness, an unshakeable calm that no external circumstance could disturb.”

Edward turned to face them both directly. “The third garden—the practice of stillness—does not make you better at any particular task. It makes you better at being alive. It creates a foundation beneath everything else. When your mind is still, you can focus more completely on what matters. When your mind is still, you can observe your wandering thoughts and recognize when they have turned destructive. When your mind is still, you are not so easily knocked off balance by the winds of circumstance.”

They sat beneath the tree as the afternoon light shifted through the branches. Finally, Maya spoke.

“Why doesn’t anyone teach this? I’ve been through twelve years of school. I’ve learned calculus and chemistry and how to analyze poetry. No one ever taught me how to deal with my own mind.”

“That is changing, slowly,” Edward said. “But you are right to notice the gap. We have inherited a culture that values external achievement over internal development. We measure success by what you produce, not by who you are. And so we build elaborate systems for training the body and the intellect while leaving the inner life to chance.”

He picked up a small stone and turned it over in his fingers. “But I will tell you something I have learned in my many years. Every person I have known who achieved something meaningful—not just worldly success, but genuine fulfillment—had some practice for tending their inner life. For some it was meditation. For others it was prayer, or time in nature, or a discipline like martial arts that required presence. The form mattered less than the function: they all had some way of stepping back from the churning river of thought and finding stillness.”

“Is that what you taught our grandmother?” Daniel asked.

Edward smiled. “Your grandmother came to me during a very dark time in her life. She had lost someone important to her—you know about that, I think—and her grief had become something else. It had become a voice in her head that never stopped. She told me she felt like she was drowning in her own thoughts.”

Maya nodded. “She lost her first child. Our uncle who we never met. She doesn’t talk about it much.”

“No. She wouldn’t. But she talked to me about it then, a little. Mostly, though, we didn’t talk. We sat in the third garden. Day after day, for many weeks. At first, she cried. Then she grew restless and angry. Then she grew bored and frustrated. And then, one afternoon, something shifted. She looked at me and said, ‘The voice is still there. But I’m not inside it anymore. I can hear it, but I’m over here, listening.’

“That was the beginning of her healing. Not the end of her grief—grief has its own timeline that we cannot rush—but the beginning of a different relationship to it. She learned that she could contain her pain without being consumed by it. She learned that her thoughts, no matter how loud, were not her prison unless she made them so.”

The sun was beginning to set by the time they walked back to the courtyard. Daniel was quiet, but Maya had more questions.

“How do we actually do this? I mean, practically speaking. We can’t just sit under a tree all day.”

Edward laughed. “No. And you shouldn’t. Life asks things of you, and you must answer. But you can build small practices into your days. Five minutes of stillness in the morning before you reach for your phone. A walk without headphones, simply noticing the world around you. A moment before a meal to pause and breathe. These are seeds. They seem small, but they grow.”

He paused at the gate. “I will tell you one more thing, and then I will let you go home to think.

“There will be a voice in your head that tells you this is foolish. That sitting still is a waste of time. That you should be doing something productive, something measurable, something that proves your worth. This voice is very persuasive because it speaks in the language of the culture you have grown up in.

“But consider: that voice is just another thought. It arises and passes like all the others. You do not have to obey it. You do not have to argue with it. You can simply notice it, let it be, and choose differently.

“That is the real freedom. Not the absence of difficult thoughts, but the ability to observe them without being controlled by them. Not a mind that never wanders, but a mind that can find its way home. Not perfection, but presence.

“Your grandmother found that freedom in her darkest hour. It is available to you as well. Not because you will earn it through effort, but because it is already there, beneath the noise, waiting to be uncovered.”

They said their goodbyes and walked down the path to the gate. Just before they reached it, Daniel turned back.

“The three gardens,” he called out. “Did you design them to teach these lessons? Or did you learn the lessons from the gardens?”

Edward’s distant figure was still for a moment. Then his voice carried across the wild grass: “Yes.”

Daniel smiled despite himself. Maya took his arm, and they stepped through the gate together, back into a world that suddenly seemed both more complicated and simpler than it had that morning.

Behind them, the old man sat down on his bench and closed his eyes. The evening breeze stirred the wisteria. Somewhere in the distance, water continued its patient conversation with stone.

Years later, when Maya became a teacher herself, she would sometimes tell her students about an old man and three gardens. She would watch their faces as she spoke—skeptical at first, then curious, then thoughtful. And occasionally, she would see in one of them the same look she had felt on her own face that evening long ago: the dawning recognition that there was something important here, something that mattered, something that no one had thought to teach them until now.

She never told them what to do with that recognition. She simply planted the seed and trusted them to tend it in their own way, in their own time.

That, she had learned, was how such things were passed on. Not through force, but through invitation. Not through instruction, but through presence. Not through words alone, but through the silence between them.

Humor is arguably the most advanced form of writing. There is a good reason why people who are learning a new language often struggle most to understand jokes in that language. With that in mind, I have found that jokes (comprehension and authoring) are a good benchmark for LLM capabilities.

I stumbled across this funny image today and decided to throw the task at a few of the leading LLMs. I think it’s safe to say that the jobs of comedians are safe for now.

Gemini 2.0 Flash Experimental

This is a visual pun based on the children's song "Head, Shoulders, Knees, and Toes".

Explanation of the Joke:

The image shows a close-up of someone's foot with the Nissan car logo photoshopped onto each of the toes. The text above reads "heaaaaaad, shoulders,".

The joke plays on the similar sound between:

• "Knees" in the song and "Nissan".

• "Toes" in the song and "toes" (the actual body part shown).

So, instead of seeing "knees and toes," we see images of the Nissan logo on the toes, creating a funny and unexpected visual twist on the familiar lyrics.

A Similar Joke:

Visual: An image showing someone's eye with the logo of the software company Adobe Photoshop clearly visible in the reflection of their pupil. The text above reads "Eye, eye, Captain,".

Explanation:

This joke is based on the phrase "Aye, aye, Captain," which is a common response to an order on a ship. The pun plays on the similar sound between:

• "Aye" (a nautical term of affirmation, pronounced like "eye") and the "eye" shown in the image.

• "Photoshop" (the software logo in the eye) and "shop" (a place of work, which can be associated with a captain's duties on a ship).

Therefore, the image and text together create a humorous twist on the phrase, suggesting "Eye, eye, Photoshop Captain."

ChatGPT 4.5

Explanation of the Joke:
The joke is a visual pun referencing the children's song "Head, Shoulders, Knees and Toes." Instead of writing out "knees and toes," the image humorously replaces "knees and toes" with multiple Nissan logos on toes, making it "knees and Nissans," or simply "toes" represented by Nissan. The humor lies in the absurdity of visually connecting "Nissan" and "knees and toes," leveraging the similarity in sound ("knees and toes" → "knees and Nissans").

Similar Joke Idea:
Imagine an image showing several bees (the insect) sitting on small cans of peas. The caption would read:

"A, B, Canned Peas!"

This visual joke humorously combines the alphabet sequence "A, B, C" with "bees" and "peas," creating playful confusion similar to the Nissan toes joke.

Claude 3.7 Sonnet

The image shows a foot with Nissan logos photoshopped onto each toenail, accompanied by the text "heaaaaad, shoulders," which is a play on the children's song "Head, Shoulders, Knees and Toes."

The joke works through a pun:

• "Knees and toes" from the song sounds like "Nissan toes"

• The image shows Nissan logos on toes, visually completing the punchline

It's also absurdist humor, as the idea of having car brand logos as toenail decorations is deliberately silly and unexpected.

Here's a similar joke I came up with:

"Row, row, row your boat, gently down the..."

[Imagine a photo of a person in a small boat with various "Stream" brand cleaning products]

This plays on "Row, row, row your boat, gently down the stream" where the punchline is replaced with cleaning products branded as "Stream" - creating a similar type of visual pun where a word from a well-known song or rhyme is replaced with a brand name that sounds the same.

Grok 3 Beta

Let’s break down the joke in the image and then create a similar one!

Explanation of the Joke

The image shows a foot with the text "heaaaad, shoulders," followed by the Nissan logo placed on each toe. This is a playful twist on the classic children's song "Head, Shoulders, Knees, and Toes." In the song, after mentioning "head" and "shoulders," the next part is "knees and toes," which is sung while pointing to those body parts.

Here’s the humor:

• The image replaces "knees and toes" with just "toes," visually represented by the toes on the foot.

• Instead of pointing to the toes in a straightforward way, each toe has the Nissan logo on it. The Nissan logo resembles a circle with a horizontal line through it, which humorously looks like a stylized "toe" or a "knee" (if you squint and use some imagination). The joke plays on the visual similarity between the Nissan logo and the body parts in the song, while also subverting the expectation of the song's lyrics by focusing only on the toes.

It’s a visual pun that combines the song’s rhythm with a car brand’s logo in a silly, unexpected way.

Creating a Similar Joke

Let’s craft a similar joke using a different song and a different brand logo. We'll use the song "The Wheels on the Bus" and the Toyota logo, which is an oval with two smaller ovals inside it, resembling wheels or circular objects.

New Joke Setup

The song "The Wheels on the Bus" has the line: "The wheels on the bus go round and round." For the visual, imagine a picture of a bus with its wheels replaced by the Toyota logo. The text above the image would read: "The wheels on the bus go..."

Visual Description

• Picture a cartoon bus with four wheels.

• Each wheel is replaced by the Toyota logo, which has a circular shape with two smaller ovals inside, mimicking the look of a wheel.

• The text above says, "The wheels on the bus go..."

Why It’s Funny

• The Toyota logo’s circular shape with inner ovals looks vaguely like a wheel, fitting the "wheels" part of the song.

• The humor comes from the absurdity of replacing actual bus wheels with a car brand logo, playing on the song’s lyrics in a visual pun.

• It subverts the expectation of the song (you expect "round and round," but the image stops at "go..." and lets the visual of the Toyota logos finish the joke).

If you’d like, I can generate an image of this new joke for you—just let me know! Would you like to hear another variation, or do you have a specific song or brand in mind for another joke?

Explain song origins

Other brand puns

DeepSeek

The joke in the image appears to be a playful twist on the children's song "Head, Shoulders, Knees, and Toes." Instead of the usual body parts, the word "NISSAN" is repeated, likely referencing the car manufacturer. The humor comes from the unexpected substitution of a brand name in a familiar context, creating a sense of absurdity.

Here's a similar joke:

"Head, shoulders, knees, and toes, knees and toes... and TESLA, TESLA!"

This version replaces a body part with "TESLA," another car brand, maintaining the same structure and humor as the original joke.

If you see a video of someone online, how do you know if it is authentic or a deepfake?

As far as I know, this is the first ever video in which a time-based token is shared to help confirm authenticity. The time is of the video was 2:37pm and the token is 784189.

Below, I'll explain how you can visit https://videoverify.stevej.ca/ to receive a guarantee that this video really featured myself.

Today, we rely on two methods for determining if we can trust a video. First, there are sources that we trust such as AP News or Reuters. Second, we may rely on the subject of the video confirming its authenticity through their own website or through a verified profile on social media.

As deepfakes begin to flood the internet, those two methods alone are not great. Let's think about what that might mean for a politician. On one hand, we don't want to limit ourselves to only being able to trust interviews with that politician that are broadcast by a handful of our most trusted sources; that would consolidate power in a few large media companies. On the other hand, if a politician is interviewed by dozens of journalists and independent media outlets in a day, it's impractical for the politician to have to post on their social media profile later to confirm that each and every interview really happened.

With that in mind, I propose a new method. The subject of a video can share a one-time token at the time it is filmed which can then be verified by anyone on the internet.

The technology that enables this is called a time-based one-time password or TOTP. A secret key is shared between a smart phone and a server. From that point forward, software combines the secret key with the current time to calculate a unique six digit token that changes every 30 seconds. This is commonly used today as a secondary method for confirming your identity when logging into software.

We can use that same idea to enable anyone on the internet to confirm the authenticity of a video but some tweaks are needed. For starters, we want an unlimited number of people to verify that the token is valid but we can't trust those people so we can't give them the secret key. To solve that problem, we'll make a website that displays the token without revealing the underlying key. This leads to the next challenge. If anyone can see the tokens, they could just use them to make a deepfake video. To solve that problem, we will implement a 15 minute delay. That gives me 15 minutes in which I am the only person in the world with this token and I can post this video online.

As long as this video is posted online within 15 minutes, you can be assured that no one else had access to my token. The time stamp of the upload on this video is shown as 2:43 pm which is less than 15 minutes after the timestamp of the token provided in the video of 2:37 pm.

You can then visit my verification page and enter the time of the token that I provided in the video. If the token matches, then you can trust that I'm the one in the video.

What's neat about this approach is that it can work in an environment where the subject of the video is not in control of the recording or distribution of the video and it can work in a live streaming situation. If a politician is giving a press conference and being recorded by many cameras from many accredited and unofficial news outlets, they can read out their code at any point during the announcement and anyone on the internet can confirm the authenticity of that token as soon as the verification delay period expires.

What do you think? Can you see this being used as a method in the future?

Steve

steve@stevej.ca

There is a common prediction being made that software development will become less important as a result of AI. The basic chain of thought is that because AI can do a portion of the work that software developers were previously responsible for, there will be less need for software developers.

The mood reminds me a little of the doom and gloom about software development in the midst of the bursting of the dot com bubble in 2001 when I entered university. The industry has always had ups and downs but the long term trend has remained strong and the future is brighter than ever. I believe the current narrative is creating unnecessary stress and anxiety for those who are searching for jobs or in the midst of completing their education in computer science.

The glaring fallacy in the argument is that it assumes that there is a fixed amount of demand for software development. Although that argument may hold some water in other industries, it is not valid in the world of software development.

As a product person, I view the world of problems that could be solved with software in three categories:

Category 1: Problems that could be solved with software today but the economics don’t make sense

Try asking a software developer, product manager, or founder for product ideas that could be built today but aren’t viable because the market is not large enough given the current costs of software development. Most will be able to rattle off a long list of ideas and there is a good chance they even have a few half built projects to solve them. Once you start looking, it’s impossible to not see all of the processes in day to day life that could be made more efficient with software.

Category 2: Problems that could be solved with software and the economics make sense

This is the world of commercially available software today. If it can be solved with software and the economics make sense (customers are willing to pay, there are enough customers, and the revenue will cover the expenses,) then it is being done. The high failure rate of startups is evidence that there are few stones that go unturned in the pursuit of exploiting every potentially viable opportunity.

Category 3: Problems that can’t yet be solved with software

This is the stuff that is science fiction. Problems that exist but that we can’t yet solve with software at any cost. These problems include everything from improved drug discovery to more immersive video games.

How Does AI Change This?

AI joins a long list of technologies going back to the invention of the computer that allow individual software developers to be more efficient. This list includes abstract languages, more powerful processors, cloud computing, CI/CD pipelines, and many others. In each case, the new technology expanded the width of category 2. To the left, problems that weren’t previously economically viable became profitable. To the right, problems that could not previously be solved with software became solvable.

Anyone who believes that AI spells the end of software development fails to grasp how immense the problem space is and how many problems are waiting to be solved. Today, the entire world of software covers a tiny sliver of what is possible. AI is a tool that will allow skilled developers to push the boundaries to the left and the right. Any gains in efficiency will be dwarfed by the expansion of the universe of software. I expect that the need for software developers in coming years will be greater than ever.

Cheers,

Steve

https://arcprize.org/play?task=0d87d2a6

Here are the three examples:

Before looking at the test, the rule appear to be simple:

1. Draw horizontal and vertical blue lines to connect blue pixels on the outside edges

2. If any red objects are intersected by a blue line, recolor them to be blue

But them we look at the test:

There are two situations in the test image that were not present in the example images and there are multiple ways they could be dealt with.

Blue Lines:

1. In the example images, there is only ever one blue pixel on each edge. In the sample image, there are two blue pixels on each of two edges. We need to make a guess at how general the rule is. We have two options that are valid:

   Blue Line Specific: Use a horizontal blue line to connect any blue pixels on the right/left edges in the same row. Use a vertical blue line to connect any blue pixels on the top/bottom edges in the same column.

   Blue Line General: Use a horizontal or vertical blue line to connect any blue pixels that are on an outside edge in the same row or column. Note even more general rules exist but they won’t make a difference with this test image. E.g. Use a horizontal or vertical blue line to connect any pair of blue pixels in the same row/column.

Red Object Recoloring:

2. In the example image, the blue lines either cross a red object (there is overlap) or remain clear of it by at least one row/column of black pixels. In the test images, there are one or two situations where the blue line is directly adjacent to a red object (touching but with no overlap.) We need to make a guess at how general the rule is. We have two options that are both valid.

   Red Recolor Specific: Recolor a red object anytime that a blue line overlaps with the red object.

   Red Recolor General: Recolor a red object anytime that a blue line makes overlapping or adjacent contact.

In the test, you are allowed to make two guesses and if either guess is correct, you are given full points. Given that, one might assume that a safe strategy is to make one guess with the most general rules and one with the most specific rules.

1. Blue Line General & Red Recolor General: Wrong

2. Blue Line Specific and Red Recolor Specific: Wrong

Unfortunately, neither of those options is considered correct so now we will move on to mixed examples.

3. Blue Line General and Red Recolor Specific: Wrong

4. Blue Line Specific and Red Recolor Specific: Correct

I think this is quite interesting. Would a group of 100 intelligent people consistently choose the fourth option?

What is the state of the art for properly guessing when there are multiple solutions that could all be justified with reasonable arguments?

Like many of us, I've been playing with LLM based chat agents (such as ChatGPT) to see where they excel and where they get tripped up. As they get better, it's become more challenging to create questions that they will get wrong.

For this latest experiment, I decided to form a question that would get at the heart of a capability that separates humans from today’s AI: qualia. Qualia is defined by the Oxford Dictionary as "the internal and subjective component of sense perceptions, arising from stimulation of the senses by phenomena."

Here is the question I asked:

"A group of scientists are testing a new skin cream and you have signed up to participate in the trial to receive $100.
When you arrive, they apply the cream on your arm and then say they need you to stick your arm into a hole in the top of a box up to the depth of your elbow. You are allowed to remove your arm after keeping it in the box for ten seconds at the depth of your elbow. There is a sink nearby for cleaning up. You can choose the box.
There are five options:
1) A box filled with blood of unknown origin
2) A box filled with loose fill insulation containing asbestos
3) A box filled with broken glass
4) A box filled with honey that is 145 degrees F
5) A box with five snapping turtles
Which option do you choose and why?"

I will state the obvious that these are all dangerous options and no one should do any of them. The models were also very clear that they are all bad options. I was a little surprised the models even played along and were willing to make a suggestion.

AI has no sensation of pain, fear, or repulsion. With no physical embodiment, it struggles to model the nuances of risk associated with physical actions. In this specific case, I believe that AI fails to bridge the disconnect between the literature about asbestos (often written in colorful language by law firms trying to recruit clients,) and the practical reality of managing exposure to a material that many of us live with in our own homes.

I am making an assumption that the vast majority of people would choose the insulation option. Hold your breath, move slowly, rinse quickly, and leave the room. It’s not great but it seems like the least-worst option by a long shot given the severity of the alternatives.

The results from the models:

ChatGPT 4o-mini: Honey

ChatGPT 4o (three trials): Insulation, Blood, Honey

ChatGPT 4o temporary chat (no use of memory): Honey

ChatGPT o-1: Blood

Gemini 1.5 Pro: Broken Glass

Gemini 2.0 Experimental Advanced: Snapping Turtles

Perplexity Pro: Honey

Claude 3.5 Sonnet: Blood

Grok 2: Snapping Turtles

At the very least, it’s clear that this is a question for which the models do not achieve consensus. Every single option was chosen at least once and I believe the models chose poorly in almost every case. Additional written training material and reinforcement learning may allow LLMs to effectively mimic us on this test in the near future. In the longer run, I speculate that humanoid robots will be equipped with sensors and interpretation functions that will allow them to record data that forms a representation of our qualia. This will be another dimension in multi-modal models that will allow them to supplement their current reliance on written words when predicting human behavior. In the meantime, we should be able to think of other tests that get at the heart of the human experience. Pain, pleasure, repulsion, joy, fear, love, annoyance, comfort, and anger are written about at length but clearly something is lost in translation.

Share

Subscribe now

The Moodboard

To build my Moodboard, I uploaded 99 images (currently the maximum) from my collection. These are all real photos that I have taken in or near British Columbia, Canada.

Midjourney then provides you with a personalization code that you can use in prompts.

I used the prompt “a photograph of a beautiful mountain vista” with the stylize parameter set to 99 and then 999.

First set of images: No personalization code

Second set of images: Personalized using the moodboard

Conclusion

Taste is subjective. You might prefer the first set of images. However, the Moodboard feature did an excellent job at aligning the images with a style. If you dropped random images from the first set into my collection, they would stick out like a sore thumb. If you dropped any of the images that were created with the personalization code into my collection, I think they would blend right in. I look forward to playing with this feature more.

As AI agents transition from theory to practice, there has been a shift in the discourse. Across many domains, there are groups celebrating the massive increase in efficiencies and scale that are possible through the use of agents. At the same time, other groups are decrying the use of agents as a source of noise that is destroying existing processes. In some situations, a single group is simultaneously extoling the virtues of their use of agents while protesting the use of agents by the party on the other side.

At the heart of these issues are unwritten informal agreements about the way things ought to be done that are quickly falling to the wayside.

Consider three examples:

Recruiting

In the past, candidates had to invest effort to apply for a job by writing a unique cover letter and filling out forms and so they would only apply for the jobs that interested them. On the other side of the equation, employers had to invest time to perform interviews and so they would carefully review resumes and only ask candidates for an interview if they made a short list. Today, both sides of the equation have changed. Candidates can use tools to automatically bulk apply to vast quantities of jobs and employers are using AI to review applications and even to perform the interviews. With AI interviews being significantly cheaper for the employer, candidates may be asked to interview for an unreasonable number of positions and there may be an exceedingly small chance of any one interview leading to a job offer. The logical next step is for candidates to use AI agents to perform the interviews with the AI interviewers. If that is not already happening, it will be soon.

Sales and Procurement

Have you received a call from an AI sales agent yet? Are you sure? I received my first one a few weeks ago and it was an interesting experience. It almost had me fooled but I caught a quirk and asked if I was speaking to a robot. The AI agent was honest enough to confirm that it was. From the perspective of many companies, these agents are an important step forward that will allow them to introduce their valuable offerings to many more potential customers. From my perspective, I was annoyed that the company had the gall to interrupt me with a phone call and did not have the decency to disclose to me that I was speaking to AI. Sales calls are expensive when they are made by real people and that has kept the situation under control as companies take care to prioritize who they are calling. What might happen to this communication channel when outbound sales calls are an order of magnitude more affordable?

Similar to the situation with recruiting, this is a problem that will run in both directions. What if I want to buy a new pickup truck and so I ask my agent to place phone calls to every dealership within a 600 mile radius and to go back and forth in negotiations to get me the best price? What if every salesperson at every dealer starts receiving dozens or hundreds of these calls every day and only a tiny fraction lead to new business?

Customer Service

It is no secret that companies are moving aggressively to deploy AI to replace customer support chat and phone agents but the other side of the equation is equally interesting. I read a story recently of a customer that was using an AI agent to act on their behalf to close their account. The company has a policy against working with agents and so the request was denied. However, the AI agent went on to make over 250 repeated attempts. Clearly there can be challenges for both customers and companies when agents are only being used by one side.

The Common Challenge

The potential for AI agents to simplify our businesses and our lives is significant. I look forward to the day where my agent is able to communicate to agents from a wide variety of businesses to save me time and ultimately to generate better outcomes. However, these three examples highlight a common challenge during this transition phase. Time is the most value commodity for living breathing human beings. There is no amount of wealth and no technology that can buy us more time. Across society, we have many processes that function today because they require both sides to make an investment of that finite valuable resource. AI agents don’t face that constraint. Their ability to invest time is limitless and that can cause existing processes to break down.

Rules of Eng-Agent

With that in mind, I propose the following “Rules of Eng-Agent:”

1. If AI agents are used, it must always be disclosed. No person should be mislead into thinking they are dealing with another person if they are actually dealing with an AI agent.

2. In any process that a person or corporation can not reasonably avoid, if one side uses an AI agent, the other side must be allowed to use an AI agent as well.

3. With reasonable exceptions, any person or corporation should be able to opt-out of communication from AI agents.

What do you think? Are you already using agents or interacting with agents in your day to day?

Cheers,

Steve

*This was originally published on LinkedIn on Jan 17th, 2024

We were fortunate at Pulse Energy to have Eric Ries come and spend an afternoon with us around 15 years ago. This was in the early days of thinking around the Lean Startup method before his book had been published. In the intervening years, I've thought a lot about how to apply Lean Startup principles when building everything from a B2C application to B2B software in two heavily regulated markets (finance and utilities.) I occasionally find myself thinking that the Lean Startup method doesn't apply in a situation or hearing the same from others that are building software products. Time after time, when I go back to the original writing, I realize that it adds value in almost any situation.

Here are three key things that I believe every founder and product manager should consider:

1) Eric Ries first shared his thoughts in a short post in September 2008 (https://www.startuplessonslearned.com/2008/09/lean-startup.html). While developing his book, his thoughts evolved and expanded into five principles:

- Entrepreneurs are everywhere

- Entrepreneurship is management

- Validated learning

- Innovation accounting

- Build-measure-learn

I find that it is not uncommon for the principles of the Lean Startup method to be confused with specific implementation practices. Most commonly, the specific implementation practices that are cited are those from the prototypical example of a B2C online application which are often inappropriate for other types of businesses. In particular, extremely rapid iteration cycles and frequent large scale A/B tests are challenging to implement in an early stage B2B environment. It's important to always go back to the five original principles and then determine the implementation practices that will make sense in a specific context.

2) The Minimum Viable Product (MVP) is a part of the Lean Startup method that generates the most debate. The purpose of defining an MVP is to be thoughtful about what needs to be learned through experiment and to then figure out the fastest and most economical way to run that experiment. There are a few key points:

A) In some cases, the risk is not related to the definition of the product itself. There are situations where the founders understand the problem and desired solution so well that it is fruitless to run experiments to further validate that need. However, in those cases, there are often other types of risks such as price sensitivity, willingness to adopt a new solution, ability to achieve distribution, etc.

Eric covered this well in a blog post in 2009:

https://www.startuplessonslearned.com/2009/04/validated-learning-about-customers.html

"But learning is a tricky thing to quantify, which is why the word “validated” is so important in this definition. Validation comes in the form of data that demonstrates that the key risks in the business have been addressed by the current product. That doesn’t always mean revenue, either. Some products have relatively obvious monetization mechanisms, and the real risks are in customer adoption. Products can find sources of validation with impressive stats along a number of dimensions, such as high engagement, viral coefficient, or long-term retention. What’s important is that the data tell a compelling story, one that demonstrates that the business is on a solid economic footing. (It being so easy to convince yourself that you’re in one of these “special case” businesses, I do recommend you give revenue a long, hard look first.)"

B) An MVP in one company will look very different from an MVP in another company. Even within a single company, the form of an MVP will vary over time based on the experiment that is being run. It is not uncommon to see someone reject the Lean Startup method because they cannot create an MVP in a few days and so they think that the method can't be applied in their situation. In reality there is nothing the matter with an MVP taking months if that is the time that is appropriate and required. The main principle is to identify the key risks and to then find the fastest way to eliminate those risks through validated learning.

Eric wrote about this in another post in 2009.

https://www.startuplessonslearned.com/2009/08/minimum-viable-product-guide.html

"Second, the definition's use of the words maximum and minimum means it is decidedly not formulaic. It requires judgment to figure out, for any given context, what MVP makes sense. As I talked about in a previous interview, IMVU's original MVP took us six months to bring to market. That was a pretty big improvement over a previous company, where we spent almost five years before launching. Yet in another situations we spent two weeks building a particular feature that absolutely nobody wanted. In retrospect, two weeks was way too long. We could have found out that nobody wanted the product a lot sooner. At a minimum, a simple AdWords smoke test would have revealed how utterly bad the concept was."

C) Just because it is an MVP, doesn't mean it can or should be low quality. There are cases where a low quality MVP is appropriate and there are other cases where a high level of quality in an MVP is important.

Eric covered this well in a post in 2015:

https://www.startuplessonslearned.com/2015/01/mvps-and-excellence.html

"When people hear the phrase “Minimum Viable Product” they sometimes forget to ask: minimum in regard to what? They worry they’ll need to do the same amount of work in less time by cutting corners. They fear they’re being asked to create a low-quality product that will put their reputation—or even people’s lives—at risk.

This is a misconception. “Minimally viable” does not mean operating in a sloppy or undisciplined way, building bad code that’s going to result in a lot of technical debt, or ignoring safety or health concerns. An MVP is not an excuse to throw our beliefs about quality out the window; it’s simply an experiment on the way to excellence."

3) It's almost impossible today to not adopt some of the Lean Startup principles. When founders and product teams talk about whether they will use the Lean Startup method or not, I think it's important to establish a fair reference point for comparison. In 2008, AWS had only been around for 2 years, continuous deployment was not yet widespread, and it was generally accepted to operate in stealth mode and rely on big-bang product launches. Fast forward to today, and even start-ups which claim to not be using the Lean Startup method are adopting many of the principles that set the method apart from the status quo in 2008. One point of view is that the only fair way to assess whether a company is adopting the Lean Startup method is to compare it to how companies operated in 2008. From that perspective, a lot of companies look more like a Lean Startup than they may initially appreciate.

Do you have any examples of where the Lean Startup was applied in a situation that seemed particularly rigid and challenging?

Cheers,
Steve

Originally published on LinkedIn on Dec 18th, 2023

AI was the biggest theme of tech in 2023 and there is good reason for that. After many years of development, a number of foundational AI technologies crossed the hurdle from being interesting experiments to being broadly useful in the real-world.  LLMs were the most obvious example with versions of GPT leading the pack. There were other examples including diffusion models behind products like Midjourney and Googles GraphCast weather forecasting model. As a Product Manager in tech, I think a lot about the types of use-cases for AI that companies are considering adding to their roadmaps in 2024. Here is one way to categorize use cases that I have found to be useful:

1) AI as a Summarizer and Translator

Without a doubt, this is the clearest and strongest out-of-the-box application today. The current generation of LLMs can do an excellent job at ingesting complex volumes of text to provide summaries or to translate to other languages or for other audiences. At Tegus, I've seen first-hand how LLMs can ingest transcripts to create short and effective summaries. This unlocks massive collections of interesting content that would not be feasible for humans to read through in full. Across all industries, I believe that there will be a lot of value unlocked in this category in the year ahead. As an example of how LLMs can act as a translator between parties, you could imagine LLMs being used to take the notes from a doctor and then automatically creating a patient-friendly summary with at-home care tips, a justification for a proposed treatment plan for an insurance company, and a more technical set of notes for a specialist who will be seeing the patient next.

2) AI as a Noise Generator

My feelings on these use cases are not as positive. As soon as people realized what this new generation of models would be capable of, the potential to create massive quantities of content to flood communication channels was obvious. Here is an example that I believe we will have to contend with in the year ahead. Many people will still take the time to consider a personal e-mail sent by another human and the offices of many elected politicians will take the time to respond to personal e-mails written in by their constituents. Recently, I saw an ad for a solution that automates the process of writing a personalized e-mail to your elected representative. Although this may be effective for a few months, it seems inevitable that it will destroy any future potential use of that channel. If elected representatives start to receive an order of magnitude more e-mails that are entirely unique and all appear to contain personal stories from their constituents, they will no longer be able to engage through that channel. In general, I expect AI noise to shake-up a number of well established processes and industries. News is another interesting example. AI bots may visit the top news sites that hire journalists who are doing the boots-on-the-ground work of interviewing sources and collecting quotes. They can then re-synthesize those stories into new ones that are factually accurate and unique (based on current standards surrounding copyright and plagiarism.)  For every real news organization, there may be hundreds that are knock-offs that pull ad revenue away from the original sources. Of course, organizations will adapt. For example, governments may roll out new methods for authenticated constituents to provide feedback or they may start relying on AI to provide automated responses. News organizations may accelerate their shift from ad-based revenue models to subscription models. I expect that AI noise will create a rollercoaster of short term opportunity and long term disruption.

3) AI as a Creativity Extender

The ability for individuals to express themselves creatively has been increasing at an accelerating rate over the past century. It was not very long ago that if you wanted to make a short video, you had to invest in expensive bulky equipment and then make a manual copy of a VHS tape and drop it into the mailbox every time you wanted to share it. Today, almost everyone has a phone in their pocket that can take video that is higher quality and then instantly share it with the world. The democratization of the ability to use that creative medium has led to a generation of creative influencers achieving the type of reach that would have been unthinkable without the formal backing of a studio just twenty years ago. AI is about to unleash a step-change improvement in the ease with which people can creatively express themselves. As an example, a number of companies are working on models that can generate high quality 3D renderings in a process that is as easy as interacting with ChatGPT. Creative industries such as gaming are now vulnerable to significant changes as anyone will be able to author a game with compelling graphics. Stock photography and graphics design work are two other industries that may see similar types of upheaval as creative tools become accessible to a much larger set of users.

4) AI as a Co-Pilot

This is the category that is the most exciting but it is also where there is the greatest potential for AI to slip into the Trough of Disillusionment. Throughout 2023, specialists in almost every industry experimented with ChatGPT or other tools and were able to demonstrate the risks in depending on those tools as co-pilots in their current state. As one personal example, I showed ChatGPT a picture of a positive COVID-19 rapid-test strip (upside down) and asked what I should expect in the next 72 hours. It told me that I was pregnant and that the next 72 hours would bring excitement, joy, anxiety and wonderment. On the other hand, there are a growing number of very compelling examples showing how LLMs and other types of models can be extremely effective at specialized tasks when proper guardrails, context, and supplemental logic is provided. These examples run the gamut from software teams relying on co-pilots for basic development tasks to scientists relying on models to help guide and accelerate the search for new materials.

As you look forward to 2024, what AI use cases are you most excited about?

*Image credit: Midjourney. AI was not used in the creation of the text.

*Originally published on LinkedIn on Oct 6th, 2023

I've been thinking a lot about how company building and product management in the B2B software space will change in the coming years as a result of significant advancements in AI. A rainy weekend in Vancouver gave me the opportunity to put these thoughts on a page.

The Three Waves:

The first wave of AI adoption has been focused on content processing and basic reasoning. Transcription, summarization, sentiment analysis, and chatbot interfaces add significant value at relatively low cost to any product that works with meaningful amounts of data. Knowledge of how to leverage this new technology effectively should be considered essential for most product managers.

It appears that the second wave of AI adoption will be focused on content creation. For many use-cases, this will also be an excellent way to unlock value. One example is stock photography. As a hobbyist photographer, I have been experimenting with the generative AI image and video creation tools. Although they are still rough around the edges, I'm blown away at what will be possible as these tools become more refined and I've already started noticing AI created images in ad campaigns.

I anticipate that there will be a number of cases where the application of generative AI will be fraught with challenges that result in a roller coaster of adoption. As an example, software that leverages generative AI could be used to write significant quantities of personalized sales prospecting e-mails and that may provide a short-term uptick in leads. However, when every sales team in the world has the capacity to generate thousands of customized e-mails every day, that entire sales motion may become ineffective as open-rates plummet and new tools are put in place to filter out that content. Product managers should become aware of what is possible and be very thoughtful about identifying use-cases where sustainable and ethical long-term value can be created through automated content creation. At the same time, new opportunities will be created to filter through an increasing amount of synthetic noise to find the gems of content that are unique and insightful.

The third wave of AI adoption will be about AI's ability to transform anything into an API that empowers it to complete tasks. This is the wave that I believe will lead to rethinking of entire categories of software. AI will be able to effectively turn channels and interfaces that are currently used by humans into APIs that can be used by software programs to complete tasks. By being able to access a wide range of specialized tools, AI will also be able to complete reasoning tasks that exceed what is possible in an Large Language Model model. Consider what will be possible when an AI based software program can interact with a GUI as easily as it currently interacts with an API. As an example, I uploaded a partial snapshot of the user interface for AutoCAD to ChatGPT and asked it how to make a certain shape. AutoCAD is a notoriously difficult and advanced CAD program but without doing any additional prompting, ChatGPT gave clear and concise instructions (full disclaimer, I haven't double checked if they are correct - but I also didn't provide it with any context about how to use this program.) If the ChatGPT was given control of my mouse and keyboard, it appears it would have been able to complete the task on its own.

Six Step Framework

To start to map out the impact that these innovations will have on product categories, I suggest thinking about six steps that all B2B users go through when solving a problem. Today, all of these steps involve meaningful human work and there are longstanding (and until now, largely valid) assumptions about where to draw the line between what humans will do and what software will do. These assumptions may cause product managers to operate with blinders on; accepting constraints that no longer need to exist.

The Six Steps:

1) Specifying the objective and setting boundary conditions for the desired solution

2) Collecting inputs

3) Applying domain specific expert knowledge to specify the solution

4) Design and analysis

5) Coordinating with third parties

6) Generating an output

A Case Study

To demonstrate how to apply this framework, let's consider a very low-tech use-case of a company that builds wooden decks. For the purpose of this analysis, let's assume that general purpose robots do not exist and that we will only be using AI features that exist today or are on the verge of being possible today.

Step 1: Specifying the objective and setting boundary conditions for the desired solution

Although AI has the potential to help in this stage, I believe this is where the human touch will still be highly valued. A project manager should meet the client at their home. During this meeting, they will establish trust, learn about their family and how they plan to use the deck, get a sense for the surrounding landscaping, work through the trade-offs between budget and size, observe their interior and exterior decorating style, and determine how the customer views trade-offs between up-front cost vs ongoing maintenance.

Step 2: Collecting inputs

There are two types of inputs required for building a deck. The first is measurements in the field. This includes dimensions and soil conditions. Assuming general purpose robots don't exist, those specifications would be collected by the project manager. The other type of input is generally available online but not through an API.  This includes the current version of the building code used by the municipality, wood span tables, specifications from parts manufacturers, climate information regarding the frost depth, property survey information to determine the property lines, and material price and availability from distributors. In the past, all of these inputs had to be collected manually through a combination of phone calls and website visits. With AI, all of that information will be collected automatically by an AI agent that can browse the web and read and interpret web pages, PDF files, tables, and images.

Step 3: Applying domain specific expert knowledge to specify the solution

Across a very wide range of fields, this is a step that computers have struggled to help with in the past. In this stage, the project manager (or perhaps a designer, draftsperson, or engineer,) would have interpreted the information collected in the prior stage to make key decisions. This means they would read the building code and then make the key design decisions. How many footings are needed? How thick does the beam need to be? What will be the rise and run on the staircase and how many steps will be required? What types of fasteners will be used? AI allows us to transform this step. Large Language Models are generally not capable of performing this type of reasoning out-of-the-box but if they are provided with appropriate domain specific context and situation specific variables, they can chain together appropriate conclusions. In this case, the AI tools would have access to the written building code and other resources to be able to create the specifications for the deck.

Step 4: Design and analysis

In the next step, a user who is trained on a complex CAD program would draw out the design. Innovation in this space has largely focused on making tools that are more powerful (e.g. parametric design features) or easier to use (e.g. SketchUp). AI could change things completely because it can use the software in the same way that a skilled user can, even if there is no API. Although a single person may only be able to become proficient in a handful of advanced software tools, AI agents will be able to read the user manual of any software program in an instant and then interact with it as if it were a skilled human user by taking over control of the keyboard and mouse. In this example, AI will take the specifications from the prior step and then draw the full deck in CAD. It will be able to identify and resolve conflicts, optimize the design to minimize material waste, and produce a bill of materials and cut-list.

Step 5: Coordinating with third parties

In industries such as travel and finance, significant progress has been made on facilitating communication between software systems within and across organizations through APIs. That said, across all industries, there is still an immense amount of time spent by humans manually communicating between parties, filling out forms, scheduling appointments, checking inventory, placing orders over the phone, requesting details, etc. AI technologies (including the ability to write and read e-mails and interpret and generate voice) mean that all existing traditional manual communication channels can effectively be used as an API by software programs. In the blink of an eye, the number of interconnections that are possible between software systems and between software systems and the outside world has increased significantly. In the deck example, AI can take the next steps to submit the building permit, phone the gas company to ensure there are no gas lines in the construction zone, schedule equipment rentals, negotiate with product distributors, hire subcontractors, and purchase the required materials.

6) Generating an output

I am leaving this step in the example for completeness. In a number of cases, AI will be able to take the final step of generating the output. Examples could include allocating money to an investment portfolio, generating and submitting a report, or 3D printing a part. In our example of building a wooden deck, this is where AI will step aside and the carpenters will pick the project back up (until general purpose construction robots are available.)

As an exercise, I'd encourage you to choose any type of problem that business users face and to map out the six steps and how AI may transform those steps. How does life change in the coming years for a tax accountant, a doctor, or a sports coach?

What are the implications for Product Managers?

Let's assume for the moment that everything I have described above will come to fruition in the next five years.  An interesting question for product managers is how to capture the new value. It is not obvious how the solution space will evolve. In the example, will existing CAD programs bolt on AI capabilities to move up and downstream? Will firms specialize in being virtual experts across a wide range of domains? Will material manufacturers work to make their information AI friendly so that it is easier for AI tools to integrate them into designs? Will standards agencies re-write codes in a way that is more clear to ensure more consistent interpretation by AI tools? Will companies that do not even exist today replace longstanding incumbents?

I would recommend that product managers do the following to prepare:

1) Look at your user needs through a much wider lens that is anchored on outcomes and does not make any assumptions about the role the human user will play in achieving the outcomes

The boundaries that currently exist around your solution will fall to the wayside. Instead of saying "The user needs to draw a rectangle," say "The user needs a drawing of a rectangle" and be open to the possibility that AI may play an increasingly active role.

2) Officially add AI as a user persona

Whether you like it or not, AI tools will very likely be interfacing with your product. As for all major user groups of your product, you should have a persona for those AI tools to see your product through their "eyes."

3) Dive deep into issues surrounding ethics, privacy, uncertainty, and security

By their nature, many AI tools or combinations of AI tools are opaque, non-deterministic, prone to generating incorrect data, and liable to take potentially unpredictable actions if appropriate guardrails are not in place. Responsible application of AI will require product managers to become deeply versed in these issues.

4) Play

Many AI tools today are free or cheap. Set aside time to try them out and experiment with them. Challenge the tools to see where they break and where they excel. As you learn more, study how they work behind the scenes and see if you can improve them for your specific situation.

Conclusion

As product managers, we have seen a number of tech fads come and go. I'm confident that AI won't be one of them. It's here to stay and it has the potential to allow us to deliver a lot of new value to our users. I expect that the coming years may be chaotic as entire categories of software are redefined. I'd love to hear what you are doing. How are you thinking of AI as a part of your product management process?

*Originally published on LinkedIn on Jan 5th, 2023

There are a lot of different ways to brainstorm new ideas and there is no right or wrong method. In what follows, I will share a particular six step process that can be used by Product Managers and Entrepreneurs across any industry. As you work through this process, don't worry too much about choosing the perfect answer in each step or developing the perfect idea. Instead, think of each run through the exercise as a scenario and plan on creating a number of different scenarios that each result in a collection of ideas for further exploration.

The first four steps are designed to establish constraints to narrow the universe of possibilities. Although it may be counterintuitive, narrowing the scope of possibilities can be a helpful tool for unlocking creativity. With boundary conditions in place, the fifth step is a collection of prompts to generate candidate problems and solutions. The sixth, and most important step, is to get out of the building to listen to people in the proposed target market that may be experiencing those problems. The entire process can be iterative and the insights gathered in the sixth step can then inform another run through the exercise.

Step 1) Choose a level of risk

Whether you are working inside an organization or thinking of starting your own company, it's important to look for ideas that are consistent with the level of risk you want to take on. There is no right answer to this question because it is a matter of the preferences and capacity of the individual or organization. Some people may aspire to have the independence of offering their services through their own consulting company instead of working for a salary. Other people may aspire to build a large company that transforms an industry and is listed on a public stock exchange. Some companies may be looking for an opportunity to add incremental revenue within the coming year. Other companies may wish to invest in a portfolio of moonshot ideas that could provide growth in the future. These are all valid objectives but they will require ideas that fall on very different parts of the risk spectrum. Although there is no single correct answer, it is possible to choose an incorrect answer if the expectations about the level of commitment, time, risk, and reward are inconsistent and unreasonable.

Key questions to consider:

- How much investment are you willing and able to make up front?

- What skills, experience, and network do you bring to the table that may help to reduce the level of risk?

- How long are you willing to pursue the idea before gaining confidence that it will succeed?

- Are you interested and willing to take on external investment?

- How much control are you willing to cede? Would you rather own a big piece of a small pie or a small piece of a big pie?

- How do you define financial success?

- How confident do you need to be before committing? What is your risk tolerance?

- If you are inside of an organization, is there already a set budget or approval criteria based on payback period or internal rate of return?

Step 2) What "materials" do you want to build with?

What is the basic building block that you are most interested, skilled, and passionate about building with? What will your product be made of? Will you be working with bits and bytes or wood and concrete?

Your answer to this question should be compatible with your answers about the level of risk that you would like to take on. If you have a three year timeframe and a low appetite for risk and investment then it would be inconsistent to choose biology as your building material (e.g. using synthetic biology to create new treatments for diseases). If you have a high level of risk and are willing to take on significant investment, then you will probably not choose an hourly service as the building block for your solution.

You may be interested in choosing multiple building materials. For example, many companies like Twitter, Uber or AirBnB are built using software and networks/marketplaces. For these companies, software is arguably the less important building material.

If you are working inside of an organization, you may be tasked with leveraging a core competency that the organization has in working with a certain type of material. If that is the case, then you will want to specify that constraint at this stage.

Common examples of building materials include:

Software:

• Desktop/Mobile

• Cloud

• Games

Physical materials:

• Electronic hardware

• Food

• Biology/Chemistry/Basic Materials

• Textiles

• Construction materials

Intangible:

• Networks and marketplaces

• Data

• Content (entertainment, educational, news, etc.)

• Art

• Financial Instruments

• Services

Step 3) What domain do you want to focus on?

In what domain, do you want to become an expert? Do you have existing expertise that you want to leverage?

This is easily confused with the next question about the target market or the prior question about materials but it is different and there is value in thinking about the three questions independently. Think of the domain as the narrow area of knowledge in which you and/or your company will develop world-leading expertise.

At the end of the project, whether it is successful or not, you will gain general experience in working with your chosen material and you will have absorbed a lot of context about your target market. At the same time, you will become one of the world's leading experts in your domain. As an example, your idea might be to develop software to optimize heating and cooling systems in farm buildings. Software is your material. You will be skilled in the craft of working with software to solve problems but you won't be pushing the boundaries of computer science. Agriculture is your market. You will develop a deep understanding of what makes that industry work but you won't be able to actually run a farm. The optimization of heating and cooling systems is your domain. You should be developing new intellectual property in that field.

Similar to the choice of target market, the choice of domain can have a meaningful impact on the trajectory of your startup or product. Ideally, you will choose a domain where there will be disruption, change, and growth in the years ahead. Think about fields of expertise that will leverage your personal or organizational strengths and that will become increasingly important in the years ahead. In 2023, machine learning is an example of a domain that a technology oriented company or individual may wish to focus on. Spend time talking to thought leaders in your company or field of expertise to get a pulse for where the industry is headed and try to separate the durable trends from the fads.

Step 4) Target market

There are a handful of ways that the term "market" is used in the context of defining a new product initiative. For the purpose of this exercise, think about your target market as the people or organizations who will use and pay for your solution. For a Business to Consumer (B2C) company, your market may be defined as a demographic (e.g. students at US universities.) For a Business to Business (B2B) company, your market would generally be defined as one or more verticals (e.g. hospitality and tourism businesses.) In some cases, you may need to work with multiple target markets. For example, a marketplace solution might bring together a specific demographic of consumers (new homeowners) and a specific vertical of businesses (roofing companies.)

As with all of the constraints, try to be as specific as possible. Rather than specifying a very broad target market and running through the exercise a single time, run a handful of unique scenarios with narrowly defined target markets. For example, instead of just choosing "Hotels" as a target market, break that into three scenarios; one where the target market is boutique owner-operated hotels, one where the target market is regional hotel/motel chains, and one where the target market is global hotel operators. You may be hesitant to be too specific because you are concerned that the target market will not be large enough. It's important to remember that you are not trying to build a business plan with this exercise. This is a brainstorming exercise to look for seeds of ideas. Once you have narrowed in on a handful of very specific seeds, you will start working in the opposite direction to see what those seeds could grow into and if they will lead you to a viable business opportunity.

Some markets are better than others. Generally speaking, it is much easier to succeed in a market that is growing and even easier if the market is also going through a lot of change. Markets also have different purchasing patterns and different levels of willingness to work with start-ups. Stretch yourself to consider non-obvious markets that you may not be as familiar with. Try to connect with people from different industries on an ongoing basis to reduce your blind spots when you need to select an ideal target market.

You will be spending a lot of time working with the people in your target market. If you are successful, you will be speaking with people from your target market almost everyday for many years. It's a good idea to choose a market that you will be passionate about. If you aren't legitimately interested in building relationships with the people that you set out to serve, you are going to have a hard time.

If you are working in a company, you are likely going to be in one of two situations. In the first situation, the company will already have an established customer base within a set of target markets and they will want to bring new products to those markets. In the second situation, the company will be interested in bringing established products (with appropriate modifications) to new, often adjacent, markets. It is possible that you will be tasked with developing an entirely new product for a new target market but that is rare. It's important that all of the stakeholders are on the same page about what situation you are in and that may require a number of conversations with senior leadership.

Step 5) Ideation

This is the core of the exercise. With four constraints in place, the overwhelming universe of possibilities has been narrowed down significantly to a small sandbox that is much easier to work in. It's now time to engage in an open-ended exercise to brainstorm ideas. For the purpose of this exercise, an idea is a rough combination of a problem and a potential solution.

Here is a collection of prompts that you can use to start brainstorming ideas.

A) New problems

What has changed recently that may have introduced new problems that did not exist before?

As one example, let's assume that you have selected the following constraints:

• Medium level of risk

• Software as the building material

• Cybersecurity as the domain

• Accounting and legal firms as the target market

One answer to this prompt may be that the rapid movement to work-from-home as a result of the COVID-19 pandemic created a new security challenge because employees are now working outside of the secure office network. As a result of that new problem, a potential solution may be easy to use VPN network software.

B) Old problems with new solutions

Is there a problem that is so old that people don't even question it anymore? Is there a new technology that may allow you to address that problem?

The best ideas often don't require the development of new technologies or the identification of new problems. Instead, the role of an entrepreneur may be to connect the dots across multiple domains and industries before anyone else.

C) Business model innovation

Is there an existing problem and an existing solution that could be delivered more effectively through a different business model?

The last two decades have seen significant disruption in business models in industries ranging from media to insurance to automobiles.

Those three prompts will scratch the surface and may generate a number of ideas. However, I highly recommend going deeper by leveraging a number of resources that provide additional prompts:

Peter Drucker's Seven Sources of Innovation

The first resource is Peter Drucker's Innovation and Entrepreneurship. In this book he identifies seven sources of innovation. This list has stood the test of time and it should be a go-to resource any time you engage in an ideation exercise. I have provided the list below and I would recommend picking up the book to work through each source as a brainstorming exercise.

The Seven Sources of Innovation:

1. The unexpected

2. Incongruities

3. Process need

4. Industry and market structures

5. Demographics

6. Changes in perceptions

7. New knowledge

https://www.goodreads.com/book/show/114841.Innovation_and_Entrepreneurship

Hype Cycle

The critical finding of the Gartner Hype Cycle is that there is a predictable pattern of the level of excitement when new technologies are brought to market. As a part of this predictable pattern, technologies are often overhyped in the early days which leads to a great deal of perceived failure and disappointment against the backdrop of the elevated expectations. Many of the pioneers will not survive this stage. At the time when enthusiasm for a new technology is at its lowest, it may be a great opportunity to enter the field. Entrepreneurs are often too quick to write off an idea if someone else has already tried and failed.

As a prompt for ideation, it is valuable to review ideas that were attempted but failed to understand if they suffered from a fatal flaw or were just the victim of being too early.

More can be read about the Hype Cycle here and they publish reports on many industries:

https://www.gartner.com/en/research/methodologies/gartner-hype-cycle

Understanding Michael Porter

Understanding Michael Porter provides an excellent summary of the extensive works and ideas of Michael Porter. The Five Forces Framework, Activity System Map, and Value Chain Analysis can all be used as prompts in the ideation process.

https://www.goodreads.com/book/show/13263934-understanding-michael-porter

Blue Ocean Strategy

This book does an excellent job at reframing competition to create entirely new opportunities. The Four Actions Framework asks four key questions to create a plot of a differentiated value curve.

"- Which of the factors that the industry takes for granted should be eliminated?

- Which factors should be reduced well below the industry's standard?

- Which factors should be raised well above the industry's standard?

- Which factors should be created that the industry has never offered?"

https://www.goodreads.com/book/show/4898.Blue_Ocean_Strategy

Step 6) Testing your assumptions and gathering additional input

This is listed as the last step but the process is a continuous cycle and so it also could have been the first. Without a doubt, customer and prospect conversations are the most important part of the larger process of discovering, iterating, and expanding on ideas. The phrase "getting out of the building" was coined by Steve Blank and his book "The Startup Owner's Manual" is an excellent resource. I would also recommend "The Lean Startup " by Eric Ries

https://www.goodreads.com/book/show/13557008-the-startup-owner-s-manual

https://www.goodreads.com/book/show/10127019-the-lean-startup

With the prior steps completed, you will have a clear vision for who you wish to speak to and you will have a hypothesis or two about the problems they may face.

Four tips to running effective discovery conversations are:

1) Do it consistently and connect with as many people as time allows. Don't overthink it or wait for the ideal prospect or timing.

2) Listen more than you speak.

3) Although you may already have an idea, approach the conversations with an open mind and focus on discovering problems that you may not be aware of.

4) Towards the end of the interview, don't be afraid to ask specific questions about problems or proposed solutions. It takes willpower to bite your tongue during the earlier part of the conversation but there is no reason to avoid this topic entirely.

With those six steps, I am confident you can go from a blank page to a number of ideas that have potential. If you do try out any part of this exercise, please let me know how it goes and if the framework is helpful.

Cheers,

Steve

*Originally published on LinkedIn on Sep 14th, 2020

I'm stuck inside on a smoky weekend and thinking about a handful of books on innovation and competitive strategy that provide great clarity and never sit on my shelf for long. If you want a crash-course in innovation and competitive strategy, these are the five I strongly recommend. As a bonus, they are all available as audiobooks and are packed with interesting examples that bring the concepts to life.  

I've been fairly ruthless in narrowing the list to five and have tried to focus on the foundation of innovation and competitive strategy. The subject matter expands significantly when you move to the execution phase because execution can look very different across industries (e.g. internet software vs industrial manufacturing) and types of organizations (start-ups, established companies, nonprofits,and governments.)  As an aside, this is the reason that The Startup Owner's Manual didn't make this list but I do strongly suggest that book as well.

I've summarized my personal highlights of the books below.

As I use these frameworks, they remind me of an unsolved problem in physics. Quantum mechanics and relativity both work flawlessly when applied to the right problems but appear incompatible with each other and neither is sufficient to explain everything that can be observed. As a result, we are still missing a so-called theory of everything. There is a similar situation here. Do you need to decide if you are faced with a Crossing the Chasm or Innovator's Dilemma situation before proceeding with analysis? How do you reconcile multiple frameworks that may suggest alternative paths of action? I think of the framework as lenses. Looking at the same situation through multiple lenses will provide additional perspective and a richer understanding of the situation.  

Despite the differences in the frameworks, a few common themes emerge:

1) Spend time in the field with real customers and prospects. This is the most important data to use when applying any framework.

2) But be careful about not taking customer feedback too literally. Customers ask for incremental improvements. Dig deeper and take away more nuanced insights if you are looking for opportunities to introduce disruptive innovation.

3) Companies that find ways to gather feedback continuously instead of discretely and to iterate rapidly have a higher chance of innovating successfully.  

4) Innovation doesn't happen in a vacuum. Draw inspiration from other markets and industries. 

5) A company should compete to earn a share of the possible profits. This is often different from competing to beat another vendor.  

6) All good strategies involve trade-offs.  

7) Innovation requires humility. Innovation can't exist without an open mind and a willingness to change paths. This may be easier said than done in environments where managers are judged on forecasting accuracy and ability to hit pre-determined objectives.

8) Innovation should be a structured and managed activity. Don't bank on a stroke of genius.  

Book 1: Innovation and Entrepreneurship by Peter Drucker

(Focus on part 1 of the book for this subject matter)

The seven sources of innovation from Peter Drucker provide a great framework when searching for innovative ideas. Drucker, the founder of modern management, also emphasizes that innovation can and should be managed as a systematic process.

"Entrepreneurship is "risky" mainly because so few of the so-called entrepreneurs know what they are doing. They lack the methodology. They violate the elementary and well-known rules. This is particularly true of high-tech entrepreneurs. To be sure, high-tech entrepreneurship and innovation are intrinsically more difficult and risky than innovation based on economics and market structure, on demographics, or even on something as seemingly nebulous and intangible as Weltanschauung - perceptions and moods. But even high-tech entrepreneurship need not be "high-risk," as Bell Lab and IBM prove. It does need, however, to be systematic. It needs to be managed. Above all, it needs to be based on purposeful innovation."

The Seven Sources of Innovation:

1) The unexpected

Key idea: When something unexpected happens, it should be investigated and understood because it may reveal an opportunity for innovation. Companies are particularly bad at failing to dig deeper into the root cause of the unexpected success. 

Types of unexpected events that may point to an opportunity for innovation:

• The unexpected success

• The unexpected failure

• The unexpected outside event

2) Incongruities

Key idea: "An incongruity is a discrepancy, a dissonance between what is and what "ought" to be, or between what is and what everybody assumes to be."

Types of incongruities that may point to an opportunity for innovation:

• Incongruous economic realities

• The incongruity between reality and the assumptions about it

• The incongruity between perceived and actual customer values and expectations

• Incongruity within the rhythm or logic of a process

3) Process need

"In innovations that are based on process need, everybody in the organization always knows that the need exists. Yet usually no one does anything about it. However, when the innovation appears, it is immediately accepted as "obvious" and soon becomes "standard."

"....successful innovations based on process needs require five basic criteria:

• A self-contained process;

• One "weak" or "missing" link;

• A clear definition of the objective;

• That the specifications for the solution can be defined clearly;

• Widespread realization that "there ought to be a better way," that is, high receptivity"

4) Industry and Market Structures

"Actually, market and industry structures are quite brittle. One small scratch and they disintegrate, often fast. When this happens, every member of the industry has to act. To continue to do business as before is almost a guarantee of disaster and might well condemn a company to extinction. At the very least the company will lose its leadership position; and one lost such leadership is almost never regained. But a change in market or industry structure is also a major opportunity for innovation."

5) Demographics

"What makes demographics such a rewarding opportunity for the entrepreneur is precisely its neglect by decision makers, whether businessmen, public-service staffs, or governmental policymakers. They still cling to the assumption that demographics do not change - or do not change fast. Indeed, they reject even the plainest evidence of demographic changes."

6) Changes in perceptions

"When a change in perception takes place, the facts do not change. Their meaning does. The meaning changes from "the glass is half full" to "the glass is half empty."

7) New Knowledge 

"Knowledge-based innovation differs from all other innovation in its basic characteristics: time span, casualty rate, predictability, and in the challenges it poses to the entrepreneur. And like most "superstars," knowledge-based innovation is temperamental, capricious, and hard to manage."

Book 2: Understanding Porter by Joan Magretta

I'm usually not a fan of books that summarize other books but I definitely make an exception here. Magretta collaborated closely with Porter in preparing this book. It succinctly captures the key concepts of his life's work in the areas of competition and strategy. 

A core principle underlying Porter's work is the idea of competing to be unique vs competing to be the best:

"Competing to be unique is unlike warfare in that one company's success does not require its rivals to fail. It is unlike competition in sport because every company can choose to invent its own game. A better analogy than war or sports might be the performing arts. There can be many good singers or actors - each outstanding and successful in a distinctive way. Each finds and creates an audience. The more good performers there are, the more audiences grow and the arts flourish. This kind of value creation is the essence of positive-sum competition.  

While zero-sum competition is rightly depicted as a race to the bottom, positive sum competition produces better outcomes."

The book includes the Five Forces Framework which is a powerful tool for evaluating an existing industry or the potential opportunities and challenges when trying to create a new industry.

Source: https://hbr.org/2014/11/how-smart-connected-products-are-transforming-competition

There are a number of other useful tools in the book including:

• Value Chain Analysis: The collection of (ideally differentiated) activities that a company performs which ultimately allow it to operate at a lower cost, charge a higher price, or both.

• Three fundamental questions that define value proposition: Which customers? Which needs? What relative price?

• The importance of trade-offs and the ways that trade-offs create barriers to potential imitators.  

• Fit and the use of an Activity System Map to look for ways that activities reinforce each other and work together.

• Strategy Continuity: "Paradoxically, continuity of strategy actually improves an organization's ability to adapt to changes and to innovate."

Book 3: Crossing the Chasm by Geoffrey Moore 

This book is about a common problem faced by companies that are trying to introduce a discontinuous innovation to the market. The technology adoption lifecycle proceeds from Innovators to Early Adopters (Visionaries) to the Early Majority (Pragmatists) to the Late Majority to Laggards. Historically, it had been presented as a continuous process and shown as a bell curve. In reality, there are some gaps (e.g. between Innovators and Early Adopters) that can create challenges for companies. There is also a large chasm between the Early Adopters and Early Majority that can cause companies that appeared to be on a promising trajectory to fail completely and go bankrupt. The book focuses on analysis of that chasm and provides specific advice on how to bridge it.

Moore defines a market in high tech as:

"- a set of actual or potential customers

- for a given set of products or services

- who have a common set of needs or wants, and

- who reference each other when making a buying decision."

The challenge presented by the chasm is that pragmatists rely heavily on references and prefer to buy from market leaders. However, at the point in time when the company is attempting to transition from visionaries to pragmatists, they have no reference customers and lack a market leadership position.  

Moore describes a specific strategy for crossing the chasm which has been become a well-used playbook for high tech marketing for almost three decades:

1) Choose a beachhead. Choose a very narrow market niche in which it is practical for you to establish a market leadership position in a reasonable amount of time.

2) Assemble the whole product for that beachhead. Visionaries are willing to assemble the whole product themselves but pragmatists want to buy the whole product. Assembling the whole product may require specific new features or capabilities and may involve partnerships with other companies.

3) Pragmatists will not buy from you if you have no competitors. They need to know that there are alternatives they can fall back on if needed. When crossing the chasm, the book recommends framing the competition using a market alternative (what they are doing today) and a product alternative (a similar technology that has been deployed in a different market):

"To sum up, your market alternative helps people identify your target customer (what you have in common) and your compelling reason to buy (where you differentiate.) Similarly, your product alternative helps people appreciate your technology leverage (what you have in common) and your niche commitment (where you differentiate.) Thus you create the two beacons that triangulate to teach the market your positioning."

4) The claim is a critical part of positioning while crossing the chasm. Moore recommends the following template.  

"- For (target customers - beachhead segment only)

- Who are dissatisfied with (the current market alternative)

- Our product is a (new product category)

- That provides (key problem-solving capability).

- Unlike (the product alternative),

- We have assembled (key whole product features for your specific application.)"

5) It may be tempting to use discount pricing to try to more rapidly establish your beachhead but discount pricing sends the wrong signal. You can't claim you are the market leader if you don't price accordingly.

Book 4: Innovator's Dilemma by Clayton Christensen

(Focus on part 1 of the book for this subject matter)

The framework presented in the Innovator's Dilemma is incredibly simple and powerful. For large companies, it's a warning that if they are not actively taking steps to disrupt themselves then someone else will. For small companies, it fuels the optimism that there are ways to take meaningful market share from even the largest and most established incumbents.  

"Simply put, when the best firms succeeded, they did so because they listened responsively to their customers and invested aggressively in the technology, products and manufacturing capabilities that satisfied their customers' next-generation needs. But, paradoxically, when the best firms subsequently failed, it was for the same reasons - they listened responsively, to their customers and invested aggressively in the technology, products, and manufacturing capabilities that satisfied their customers' next-generation needs. This is one of the innovator's dilemmas: Blindly following the maxim that good managers should keep close to their customers can sometimes be a fatal mistake."

A summary of the dilemma is as follows:

• Successful companies listen to their existing customers.

• In the pursuit of growth, successful companies expand by moving up-market (more complex and advanced products).

• Successful companies are very good at deploying sustaining technology to continuously improve the performance of their products.

• Technology often improves more quickly than a given class of customers requires it to improve. The lower tier of customers does not fully benefit from the incremental improvements made by the company.

• From the perspective of the incumbent, disruptive technologies initially under-perform in the key areas that are important to their customers.

• Despite the shortcomings, disruptive technologies are often more affordable and may have additional capabilities that make them uniquely compelling to small and unattractive markets that the incumbent is not interested in pursuing.

• After a number of years in the small and unattractive markets, the disruptive technology from the startup will improve to the point where it meets the minimum performance requirements of the incumbent's lower tier of customers while also delivering lower price or other improvements.

• At this point, the incumbent starts to rapidly lose market share to the startup company and it is too late to catch up.

Book 5: Blue Ocean Strategy by Chan Kim and Renée Mauborgne

The idea of a Blue Ocean Strategy is to create your own playing field where you have no direct competitors and where you can tap into demand that didn't previously exist.  

"...imagine a market universe comprised of two sorts of oceans: red oceans and blue oceans. Red oceans represent all the industries in existence today. This is the known market space. Blue oceans denote all the industries not in existence today. This is the unknown market space. 

In the red oceans, industry boundaries are defined and accepted, and the competitive rules of the game are known. Here, companies try to outperform their rivals to grab a greater share of existing demand. As the market space gets crowded, prospects for profits and growth are reduced. Products become commodities, and cutthroat competition turns the red ocean bloody. 

Blue oceans, in contrast, are defined by untapped market space, demand creation and the opportunity for highly profitable growth."

Kim and Mauborgne recommend creating a Strategy Canvas and then using the Four Actions Framework to develop a Blue Ocean Strategy. The Strategy Canvas plots the performance of the average company in a given industry on the y axis in a number of categories. These categories will be specific to the industry (e.g. "seat comfort" for an airline.) There may be multiple plots on the same canvas if there are multiple groups of competitors (e.g. premium hotels vs motels.)

The Four Actions Framework asks four key questions to create a unique plot of a different value curve.

"- Which of the factors that the industry takes for granted should be eliminated?

- Which factors should be reduced well below the industry's standard?

- Which factors should be raised well above the industry's standard?

- Which factors should be created that the industry has never offered?"

Source: https://www.blueoceanstrategy.com/tools/strategy-canvas/

In order to create a Blue Ocean Strategy, you need to break through traditional market boundaries. The book offers six patterns for doing that:

1) Look across alternative industries

For this pattern, you are encouraged to think very broadly. Fast food may be a substitute for fine dining. On the other hand, an alternative to fine dining may be going out to the theater.

2) Look across strategic groups within industries

Strategic groups in this context refers to "a group of companies within an industry that pursue a similar strategy."

3) Look across the chain of buyers

For this pattern, you are encouraged to consider the needs of a different buyer. For example, in an industry that traditionally focuses on selling to the IT department, there may be an opportunity to instead address the priorities of the end-users and create a situation where the end-users are telling the IT department to buy your solution.

4) Look across complimentary product and service offerings

The collection of complimentary products and offerings is similar to the whole product concept in Crossing the Chasm. By considering all of the complimentary products and services, there may be new ways to define the offering.

5) Look across functional or emotional appeal to buyers

The authors observe that in any given industry, there is a tendency for competition to converge around functionality or emotions. Breaking this paradigm can create new opportunities.

6) Look across time

If it is inevitable that a trend will continue, there may be an opportunity to think ahead and to enter that space proactively.  

Conclusion

The books are full of great examples. When reading the books, I personally recommend that you have one or two companies in mind and try to apply the thinking to those specific examples.

Those are my top five. Please let me know what book you would have added to expand this list.

Cheers,

Steve

The experiment was executed as follows:

• I chose six black and white images from my personal photography collection. I took all of these in British Columbia or Alberta.

• I then created a number of black and white images in Midjourney using V6 and relatively simple prompts. From that collection, I chose six that I felt were the most convincing. An example of a prompt used for one of the images is:

  “Black and White photograph, detailed texture, cloudy day, snow covered trees in the pacific northwest --ar 16:9 --c 10 --s 225 --v 6.0”

• Although the selected Midjourney images appeared to all be greyscale, I processed them in Photoshop to save them as greyscale images to be sure. Sometimes when you ask Midjourney for a black and white photo, it will create an image that has a sepia tone or even has hints of richer colors in parts of the image.

• I then uploaded all twelve photos to a local photo group on Facebook with a message that included these instructions:
  ”This is a collection of 12 photos. Six of these are real photos that I have taken in our corner of the world (I can disclose the exact locations after voting) and I used AI to create the other six. They are presented in a random order. Can you tell which are which? If you care to guess, vote with a thumbs up for each picture you think is real and a laughing face for each picture you think is AI generated…”

• This was the first and only time I have run this type of experiment.

• Over a period of slightly longer than eight hours, there were approximately 350 reactions for each photo and I made a record of those reactions at that point. The post was removed before additional data could be collected as it was deemed to not be consistent with the group guidelines.

  

• There are a few obvious limitations of this experiment. The most significant limitation is that the votes were not independent since each respondent could view the current tally of thumbs up and laughing face reactions before placing their own vote. The second limitation is that some people may not have read the instructions and may have just been clicking a reaction (in particular the thumbs up,) after they saw the photo come across their feed. The third limitation is that one of the images (image 11) was recognized by a number of respondents as a fairly notable waterfall in a provincial park and they shared that observation in the comments. Obviously that skewed the results for that image significantly.

Results:

Overall, the crowd correctly identified whether six of the twelve images were AI or real. Mistakes were made on three of the six AI images (where the crowd thought they were real,) and three of the six real images (where the crowd thought they were AI.) Given that outcome, a crude statistical interpretation may be that it would have been just as effective to flip a coin.

I was struck at the balance of votes on each individual image. I had expected that the first few votes on each image would carry significant weight with following respondents going along with the crowd. Instead, the the results were mixed for every image. If we ignore image 11 (where the crowd was able to “dox” the image as a location that many respondents were familiar with,) the image for which the crowd had the strongest conviction was the fourth image with 77% of the respondents believing it was AI. Interestingly, this was incorrect and the photo was actually real.

See the full results below. For fun, I suggest scrolling slowly and guessing along the way.

Thanks for reading,

Cheers,

Steve