Quico uses a common logical fallacy called appeal to historical precedent (printing press, electricity, internet). This is a false analogy exactly because AI potentially transforms reasoning itself, the meta-capacity humans use to solve problems. AI is not a domain-specific tool even if one of its capacities allows you to use it as one. It is also called Black Swan blindness which is the error of assuming rare catastrophic events won't occur because they haven't occurred yet.
Aldous Huxley argued that the greatest threat to human freedom wasn't tyranny but comfort. That people would surrender autonomy not under force but for the sake of ease. Quico’s love of AI for the sake of convenience echoes that. There is a saying that the road to hell is paved with good intentions.
It was funny when Quico joked that he felt like a drug-pusher. And, like people who are addicted to substances, they are the last ones to understand when they are no longer in control – so Quico’s confidence that he will know when AI diminishes his autonomy is misplaced.
I am sure Quico is a nice guy, but his response that art is magical and sacred was condescending and dismissive. Quico should read the book titled Objectivity by Harvard professors of the history of science Lorraine Daston and Peter Galison to understand the line between science and the humanities is much, much thinner than he knows.
While we cannot stop what some call progress, we need more countervailing voices like Sam’s so that we can shape it to serve what is good about humans rather than what is bad.
It’s difficult to go deep via the comments, but I do want to also correct Quico’s claim that AI has actually resulted in the need for more radiologists. Once you control for two factors, AI will replace radiologists: a) boomers are flooding the medical space driving up imaging numbers, and b) AI is not sophisticated enough yet to eliminate the need for what Lorraine Daston and Peter Galison call “trained judgment” that incorporates individual context. AI is mostly pattern recognition. It cannot evaluate individual context — yet. Once you remove those factors, AI will drastically reduce the number of radiologists needed. If it develops context integration (AGI) then it will most likely eliminate them.
This just came out from MIT — everyone should read it. Unfortunately the policy solutions are non-starters. There will be no political will to carry them out.
I think many of us, even those deeply concerned about the ecological, social, economic and cognitive impacts of AI might continue to use certain relatively well established and clearly helpful kinds - like predictive journey mapping and translation services. But if I were god (a common joke in our households), I’d definitely ban it from frivolous purposes and I’d certainly insist on every request being accompanied by a planetary and financial costs bill. Classic example of what economists call externalized costs that have run the planet and climate into the ground.
Watch this documentary: The AI Doc: Or How I became an Apocaloptimist. The ending is ridiculous, but he managed to get three of the five CEOs on camera. Also, we know they are light-years beyond whatever they are willing to discuss in the current moment.
Yes, you’d have to be a God in order to limit AI to domain-specific tasks. It’s already out in the wild; there’s no going back. Now it’s about how do we go forward responsibly. Maybe that train has already left the station too? Most people are drunk on its power to make life easier. I do believe the cognitive effects will rival the climate ones in that people will no longer know much of anything for themselves. The human brain has not changed in how it processes information and learns from that information. The brain records through meaning not exposure — requires multiple modalities of processing to fully encode information. People will become AI parrots — they sound smart but will be incapable of explaining what they just said. I see it already. People I know who are otherwise intelligent are simply using AI as if it’s an all-knowing God or a type of oracle in that they do not fact-check it. One person used AI’s errors to refute actual facts. Sure, AI will get better with fewer errors, but human understanding will not. It is likely to disappear.
I would say definitely boycott AI, starting today. I will follow your progress, and if it looks like you're successful after six months, I will join your boycott with my massive subscriber list, currently numbering in the low double digits [including my mailman & amazon drivers]. I have studied the history of technological dislocation, and in every case in which citizens resisted a disruptive innovation, they successfully impeded the adoption of that innovation. Which is why I still drive a steam-powered automobile with whale-oil lantern headlights.
Cheer up, Chuck. At this rate, we’ll all be doing that again within the century - those of us who have survived the fall from the cliff. I can’t believe the dislocation from ecological and climate reality of some of the techno-optimists I meet. We are entering a realm of ‘the higher they come, the harder they fall’ illustrations. The USA is one of the first, but it won’t likely be the last.
I happen to think humans lost the ability to control our choices either in the first industrial revolution, or way back with the transition to agricultural settlements and the dawn of permanent civilization. We are stuck in two kinds of arms races: one of civilization against the forces of nature, the other with our human peers. I don't think we really have the choice to avoid innovation, no matter the cost. We only pretend to have the choice, so we feel more in control of our own future. But I guess we'll see what happens, eh?
I guess we will. But I see an amazing amount of agency in humanity right now. Will it be enough to counter the power and control-grabbing tech bros and oligarchs and idiots? That’s the question. But as my colleague Dave (Sir David King) says, they’ll have to wheel me out of here in a coffin before I stop doing my work to shift our trajectory. 😉
You just gave me an idea: Why don’t coffins have wheels, like luggage, or kids’ shoes that have pop-out wheels in the heels for skating? If I end up commercializing your innovation, I’ll keep you in mind for royalty comps. I think your coffin concept could be huge…I’mma work up an AI pitch deck.
It might be a mistake to think that billionaire tech bros are mere greedy idiots, just like it might be a mistake to think that politicians we don’t like are idiots. I would say try doing their jobs for a month before coming to a conclusion on their capabilities. There’s a difference between disliking a personality and evaluating genuine competence. If memory serves, 90% of the people who now call Elon Musk an idiot for his political migration used to call him a genius only five years ago. If someone can actually rise or fall by several orders of magnitude in IQ according to shifts in political alignment, that is actually earth-shattering news for the field of behavioral psychology, but I have not seen any white papers devoted to that subject. There is certainly a difference between MANAGERIAL ingenuity [such as Steve Jobs] and ENGINEERING ingenuity [such as Steve Wozniak], but the distinction in general intelligence is hard to assess on those metrics. I would have difficulty guessing the relative performance of Jobs v Wozniak on a random task unrelated to their fields of expertise; but Jobs is dead, so I’d expect Wozniak to do a lot better if the test were given today.
Oh don’t get me wrong. Musk is not an idiot. Deluded, yes I know so. Self-aggrandizing, maybe but in an unusual way. And I’ve worked in government for quite a long time myself so I’ve observed both gene pools at length and close range. I was really only referring to some of the Trumpites, and perhaps for some it’s a fairly pronounced mental illness rather than idiocy in the conventional sense. Nonetheless, I should be more precise, okay.
I’m a Trumpite myself. From 2008-2012, I was an Obama-ite. What does that make me, good person or bad person? Genius or moron? I can be only one of each category, so choose carefully. I will be forced to wear the label you give me for the rest of my life. Thank you for your attention to this matter!!!
Thanks Quico, Luke and Sam for this. I watched most of this - until the Q&A opened - and think you both, all, made good points.
As a climate and biodiversity scientist, my postdocs used early AI (machine learning) to improve climate models - a good example, like vaccine development or translation services, that’s a genuine public good.
But what astonishes me to be honest is the failure of most of this debate even to rise above the personal preferences to the overwhelming systemic issues. AI use, released as a default option to the entire world on search engines and to make silly memes about JD Vance or cats, is driving us over the climate stability cliff - and the water security cliff too perhaps. As a senior thinker in Anthropocene Institute, Quico, I’m disturbed that you didn’t even raise this, and nor did Luke.
AI is a classic example of what Sam did, eventually raise (in talking about Kentucky): yep, we can do this, great - but should we? Jobs, culture, erosion of what has always been a good life, mental development of those who didn’t grow up developing their skills before AI (thanks to the audience member who raised it) - all that. We are in an incredibly important crossroads of history, with 8.2 billion people on the planet. Yep, it saves time - and I agree it can be incredibly useful research. But do we really need it? Should we even believe that we want it and can’t do without it? Governance and ethics issues, as usual, lag behind the market.
Personally, I don’t think we should use it in an unrestricted way. We are restricting its use severely in our films. Ask John Bowey about this when you and we speak soon! Thanks for a stimulating debate.
I teach writing to undergraduates and I'm afraid Quico's approach to AI would spell doom for the next generation of competent writers and editors were they to adopt it. Fortunately most of my students know this and turn in messy, half baked work that we go over together in person as part of the revision process. We call it “productive inefficiency.” Sam had the better arguments here but also the better values, in my view.
Quico uses a common logical fallacy called appeal to historical precedent (printing press, electricity, internet). This is a false analogy exactly because AI potentially transforms reasoning itself, the meta-capacity humans use to solve problems. AI is not a domain-specific tool even if one of its capacities allows you to use it as one. It is also called Black Swan blindness which is the error of assuming rare catastrophic events won't occur because they haven't occurred yet.
Aldous Huxley argued that the greatest threat to human freedom wasn't tyranny but comfort. That people would surrender autonomy not under force but for the sake of ease. Quico’s love of AI for the sake of convenience echoes that. There is a saying that the road to hell is paved with good intentions.
It was funny when Quico joked that he felt like a drug-pusher. And, like people who are addicted to substances, they are the last ones to understand when they are no longer in control – so Quico’s confidence that he will know when AI diminishes his autonomy is misplaced.
I am sure Quico is a nice guy, but his response that art is magical and sacred was condescending and dismissive. Quico should read the book titled Objectivity by Harvard professors of the history of science Lorraine Daston and Peter Galison to understand the line between science and the humanities is much, much thinner than he knows.
While we cannot stop what some call progress, we need more countervailing voices like Sam’s so that we can shape it to serve what is good about humans rather than what is bad.
It’s difficult to go deep via the comments, but I do want to also correct Quico’s claim that AI has actually resulted in the need for more radiologists. Once you control for two factors, AI will replace radiologists: a) boomers are flooding the medical space driving up imaging numbers, and b) AI is not sophisticated enough yet to eliminate the need for what Lorraine Daston and Peter Galison call “trained judgment” that incorporates individual context. AI is mostly pattern recognition. It cannot evaluate individual context — yet. Once you remove those factors, AI will drastically reduce the number of radiologists needed. If it develops context integration (AGI) then it will most likely eliminate them.
This just came out from MIT — everyone should read it. Unfortunately the policy solutions are non-starters. There will be no political will to carry them out.
https://economics.mit.edu/sites/default/files/2026-02/AI,%20Human%20Cognition%20and%20Knowledge%20Collapse%2002-20-26.pdf
I think many of us, even those deeply concerned about the ecological, social, economic and cognitive impacts of AI might continue to use certain relatively well established and clearly helpful kinds - like predictive journey mapping and translation services. But if I were god (a common joke in our households), I’d definitely ban it from frivolous purposes and I’d certainly insist on every request being accompanied by a planetary and financial costs bill. Classic example of what economists call externalized costs that have run the planet and climate into the ground.
Watch this documentary: The AI Doc: Or How I became an Apocaloptimist. The ending is ridiculous, but he managed to get three of the five CEOs on camera. Also, we know they are light-years beyond whatever they are willing to discuss in the current moment.
https://m.imdb.com/title/tt39150120/?ref_=ext_shr_lnk
Yes, you’d have to be a God in order to limit AI to domain-specific tasks. It’s already out in the wild; there’s no going back. Now it’s about how do we go forward responsibly. Maybe that train has already left the station too? Most people are drunk on its power to make life easier. I do believe the cognitive effects will rival the climate ones in that people will no longer know much of anything for themselves. The human brain has not changed in how it processes information and learns from that information. The brain records through meaning not exposure — requires multiple modalities of processing to fully encode information. People will become AI parrots — they sound smart but will be incapable of explaining what they just said. I see it already. People I know who are otherwise intelligent are simply using AI as if it’s an all-knowing God or a type of oracle in that they do not fact-check it. One person used AI’s errors to refute actual facts. Sure, AI will get better with fewer errors, but human understanding will not. It is likely to disappear.
I would say definitely boycott AI, starting today. I will follow your progress, and if it looks like you're successful after six months, I will join your boycott with my massive subscriber list, currently numbering in the low double digits [including my mailman & amazon drivers]. I have studied the history of technological dislocation, and in every case in which citizens resisted a disruptive innovation, they successfully impeded the adoption of that innovation. Which is why I still drive a steam-powered automobile with whale-oil lantern headlights.
Cheer up, Chuck. At this rate, we’ll all be doing that again within the century - those of us who have survived the fall from the cliff. I can’t believe the dislocation from ecological and climate reality of some of the techno-optimists I meet. We are entering a realm of ‘the higher they come, the harder they fall’ illustrations. The USA is one of the first, but it won’t likely be the last.
I happen to think humans lost the ability to control our choices either in the first industrial revolution, or way back with the transition to agricultural settlements and the dawn of permanent civilization. We are stuck in two kinds of arms races: one of civilization against the forces of nature, the other with our human peers. I don't think we really have the choice to avoid innovation, no matter the cost. We only pretend to have the choice, so we feel more in control of our own future. But I guess we'll see what happens, eh?
I guess we will. But I see an amazing amount of agency in humanity right now. Will it be enough to counter the power and control-grabbing tech bros and oligarchs and idiots? That’s the question. But as my colleague Dave (Sir David King) says, they’ll have to wheel me out of here in a coffin before I stop doing my work to shift our trajectory. 😉
You just gave me an idea: Why don’t coffins have wheels, like luggage, or kids’ shoes that have pop-out wheels in the heels for skating? If I end up commercializing your innovation, I’ll keep you in mind for royalty comps. I think your coffin concept could be huge…I’mma work up an AI pitch deck.
It might be a mistake to think that billionaire tech bros are mere greedy idiots, just like it might be a mistake to think that politicians we don’t like are idiots. I would say try doing their jobs for a month before coming to a conclusion on their capabilities. There’s a difference between disliking a personality and evaluating genuine competence. If memory serves, 90% of the people who now call Elon Musk an idiot for his political migration used to call him a genius only five years ago. If someone can actually rise or fall by several orders of magnitude in IQ according to shifts in political alignment, that is actually earth-shattering news for the field of behavioral psychology, but I have not seen any white papers devoted to that subject. There is certainly a difference between MANAGERIAL ingenuity [such as Steve Jobs] and ENGINEERING ingenuity [such as Steve Wozniak], but the distinction in general intelligence is hard to assess on those metrics. I would have difficulty guessing the relative performance of Jobs v Wozniak on a random task unrelated to their fields of expertise; but Jobs is dead, so I’d expect Wozniak to do a lot better if the test were given today.
Oh don’t get me wrong. Musk is not an idiot. Deluded, yes I know so. Self-aggrandizing, maybe but in an unusual way. And I’ve worked in government for quite a long time myself so I’ve observed both gene pools at length and close range. I was really only referring to some of the Trumpites, and perhaps for some it’s a fairly pronounced mental illness rather than idiocy in the conventional sense. Nonetheless, I should be more precise, okay.
I’m a Trumpite myself. From 2008-2012, I was an Obama-ite. What does that make me, good person or bad person? Genius or moron? I can be only one of each category, so choose carefully. I will be forced to wear the label you give me for the rest of my life. Thank you for your attention to this matter!!!
Thanks Quico, Luke and Sam for this. I watched most of this - until the Q&A opened - and think you both, all, made good points.
As a climate and biodiversity scientist, my postdocs used early AI (machine learning) to improve climate models - a good example, like vaccine development or translation services, that’s a genuine public good.
But what astonishes me to be honest is the failure of most of this debate even to rise above the personal preferences to the overwhelming systemic issues. AI use, released as a default option to the entire world on search engines and to make silly memes about JD Vance or cats, is driving us over the climate stability cliff - and the water security cliff too perhaps. As a senior thinker in Anthropocene Institute, Quico, I’m disturbed that you didn’t even raise this, and nor did Luke.
AI is a classic example of what Sam did, eventually raise (in talking about Kentucky): yep, we can do this, great - but should we? Jobs, culture, erosion of what has always been a good life, mental development of those who didn’t grow up developing their skills before AI (thanks to the audience member who raised it) - all that. We are in an incredibly important crossroads of history, with 8.2 billion people on the planet. Yep, it saves time - and I agree it can be incredibly useful research. But do we really need it? Should we even believe that we want it and can’t do without it? Governance and ethics issues, as usual, lag behind the market.
Personally, I don’t think we should use it in an unrestricted way. We are restricting its use severely in our films. Ask John Bowey about this when you and we speak soon! Thanks for a stimulating debate.
I teach writing to undergraduates and I'm afraid Quico's approach to AI would spell doom for the next generation of competent writers and editors were they to adopt it. Fortunately most of my students know this and turn in messy, half baked work that we go over together in person as part of the revision process. We call it “productive inefficiency.” Sam had the better arguments here but also the better values, in my view.
Kahn can't argue. He's a ranter. A really, really good one. Ironically, AI will never take that job, even as it assists researching, editing, etc.