Protect cognitive skills by limiting AI synthesis usage
“I would say any skill you want to preserve in your head, you should probably not use AI for. So I use AI for editing right now. You very quickly end up on a slippery slope. There are already scientists and researchers looking at the negative cognitive impacts of depending on AI, much like your ability to navigate is probably deteriorated since using Google Maps. In some respects, each individual is more enhanced, augmented using these tools. But if you do want to keep certain muscles strong and able, that's where I would hesitate. But if you lose it, it's a hell of a lot harder to reclaim it.”
Hacking mastery now requires significantly less human effort
“Twenty-five years ago, there were a million bugs being found in the Windows operating system, and for that to happen, people had to really dig into the ins and outs of how the Internet interacted with Windows. But it required hours and hours of work for humans to achieve the level of mastery required to even be playing in the bug hunting game. AI changes all that, right? Like, AI can just look at all these bugs and kind of get to that level of mastery very quickly.”
“It's that you have a certain threshold, let's call it $35,000 at the poverty level. If you make less than that, then we true you up to that level. If you make more than that, then it's as it was. That's the thing that I'd be thrilled with. I'd be thrilled with a higher child tax credit. I'm thrilled with anything that alleviates poverty.”
Anthropic restricts model access to prevent systemic collapse
“Anthropic was talking about it as very dangerous, you know, like we're not sure what to do with this, like who should get it? They picked about 50 corporations and organizations and said, take a look at this, see what you can do with it. The idea is that access to mythos could give those companies a head start against bugmageddon, allowing them to find the holes in their systems and patch them before hackers get their hands on mythos.”
Hacking mastery now requires significantly less human effort
“Twenty-five years ago, there were a million bugs being found in the Windows operating system, and for that to happen, people had to really dig into the ins and outs of how the Internet interacted with Windows. But it required hours and hours of work for humans to achieve the level of mastery required to even be playing in the bug hunting game. AI changes all that, right? Like, AI can just look at all these bugs and kind of get to that level of mastery very quickly.”
Hacking mastery now requires significantly less human effort
“Twenty-five years ago, there were a million bugs being found in the Windows operating system, and for that to happen, people had to really dig into the ins and outs of how the Internet interacted with Windows. But it required hours and hours of work for humans to achieve the level of mastery required to even be playing in the bug hunting game. AI changes all that, right? Like, AI can just look at all these bugs and kind of get to that level of mastery very quickly.”
Cybersecurity faces a modern Y2K style software crisis
“However you slice it, it's the Y2K problem for AI. In cybersecurity, we always talk about the awful things, the ransomware outbreaks and hacks and things like that. But occasionally, we do something right collectively. And Y2K was an example of when the world knew about a problem and worked really hard to fix it.”
Resistance training builds white matter; cardio builds the hippocampus
“So you see that aerobic training or interval training seems to benefit the hippocampus and memory function. Resistance training seems to particularly benefit the white matter of the brain, which is kind of in the middle of the brain. That's where all the myelin is. That's responsible for fast connections and some of our more complex cognitive functions. Resistance training improves the structure and function of the white matter and with that improvements in things like executive function.”
AI models like Mythos find deep software vulnerabilities
“Anthropic's Mythos model is so capable that the company restricted access to 12 partners and a $100 million compute budget rather than releasing it publicly. It has already identified 20 zero-day vulnerabilities in decades-old software. Now the question over DeFi: if Mythos turns its attention to smart contracts, what survives?”
“In a world full of tool systems and AI, what human abilities or habits are becoming more valuable, not less? I would say the relational, the tactile, anything IRL in real life that can be extended also to, for instance, in my case, informational advantage, offline informational advantage. A lot of the LLMs are slicing and dicing the internet. One might argue all of them are doing that. And whether you are looking at longevity in professional terms, if you're looking at longevity in creative terms, I think putting on the lens of looking at what you can do in IRL that currently... allows me to have an informational advantage because none of that is online.”
Anthropic restricts model access to prevent systemic collapse
“Anthropic was talking about it as very dangerous, you know, like we're not sure what to do with this, like who should get it? They picked about 50 corporations and organizations and said, take a look at this, see what you can do with it. The idea is that access to mythos could give those companies a head start against bugmageddon, allowing them to find the holes in their systems and patch them before hackers get their hands on mythos.”
“Placement rates have collapsed for particular populations, particularly computer science grads coming out of various universities, where you had programs that had placement rates of 94% at high salaries. And those have flipped. Now you have computer science grads out of great programs looking for six months and getting AI-fueled responses. To me, that population is the canary in the coal mine.”
“The CEO I had breakfast with who's freaking out, he said three words to summarize the situation: 'Capital displaces labor.' And then he put in parentheses 'with the help of AI.' But he's like, now it's clearer to him than ever because his company is roaring in terms of performance and stock market price, and he's cutting people right and left.”
“And so he talks about the idea of cognitive prosthetics versus cognitive orthotics. A prosthetic is something that you use if you are missing a function. An orthotic is something you use to boost your function. And so we can use AI as orthotics. We can do the work ourselves and then, you know, say to whatever tool it is, you're like, what am I missing? What can I do better? All of this requires you to be engaged in the work to actually think yourself, and you're building your skills.”
Independent candidates can rationalize broken politics
“If you can get someone like Seth Bodnar into the US Senate, all Senate votes count the same. You could have a sane, public-spirited, patriotic senator out of a traditionally red state in Montana. And so that's the kind of candidate I'm boosting and trying to help in the goal of rationalizing our politics come January of next year.”
Cybersecurity faces a modern Y2K style software crisis
“However you slice it, it's the Y2K problem for AI. In cybersecurity, we always talk about the awful things, the ransomware outbreaks and hacks and things like that. But occasionally, we do something right collectively. And Y2K was an example of when the world knew about a problem and worked really hard to fix it.”
Exploitation speed has collapsed from years to days
“Eight years ago, the average time between a bug being found and a hacker using that bug in a cyber attack was 847 days. Now it's like within a day. It's not rocket science, but it takes time for a human to do it. You have to have a certain level of expertise. AI has absorbed all of that.”
Exploitation speed has collapsed from years to days
“Eight years ago, the average time between a bug being found and a hacker using that bug in a cyber attack was 847 days. Now it's like within a day. It's not rocket science, but it takes time for a human to do it. You have to have a certain level of expertise. AI has absorbed all of that.”
Anthropic restricts model access to prevent systemic collapse
“Anthropic was talking about it as very dangerous, you know, like we're not sure what to do with this, like who should get it? They picked about 50 corporations and organizations and said, take a look at this, see what you can do with it. The idea is that access to mythos could give those companies a head start against bugmageddon, allowing them to find the holes in their systems and patch them before hackers get their hands on mythos.”
Independent candidates can rationalize broken politics
“If you can get someone like Seth Bodnar into the US Senate, all Senate votes count the same. You could have a sane, public-spirited, patriotic senator out of a traditionally red state in Montana. And so that's the kind of candidate I'm boosting and trying to help in the goal of rationalizing our politics come January of next year.”
Cybersecurity faces a modern Y2K style software crisis
“However you slice it, it's the Y2K problem for AI. In cybersecurity, we always talk about the awful things, the ransomware outbreaks and hacks and things like that. But occasionally, we do something right collectively. And Y2K was an example of when the world knew about a problem and worked really hard to fix it.”
“But at the root, we do, in fact, go back to the fact that a single party had a single key, that thing was compromised by some bad actors, and that allowed them to take unilateral action. And apparently, this unilateral action basically, minting a whole bunch of USR was either not monitored or maybe they had alerts on it, but nobody was monitoring those alerts.”
Cultivate courage through action and progressive resistance
“I think courage is learned. You have to practice it. And if you're not afraid, it's not courage, right? If someone's fearless, they're by definition not using courage. You have to be afraid of something. So you can edge yourself and you can edge kids into that. I don't think courage is a decision. I don't think courage is something you get from reading a book. I think you have to prove to yourself that you have it. And the only way your subconscious will believe it is if you are actually doing things that are uncomfortable. It is through action and progressive resistance that you develop courage.”
Enforce zero-tolerance policies to maintain healthy communities
“Somebody walks into my house. This is a shoes-free house. Let's say somebody comes in tracking mud all over the place. That person's going to get dragged by their hair out and then they're never coming back in. Zero tolerance policy for broken windows. When these minor infractions are permitted, the Overton window, the broadness of what is now allowable behavior shifts. If you allow minor infractions, you're going to get moderate infractions. You allow those, you're going to get major infractions. From the very first days of the blog, the comments section has guidelines. If you're an asshole, we're going to boot you.”
Government-funded job retraining has largely failed
“When I looked up the studies as to how effective government-funded retraining programs were, the efficacy range I found was 0-15 percent, with 15 percent on the high side. Trying to train workers who've been laid off in various jobs to compete against AI strikes me as a loser. It strikes me as the next generation of 'learn to code' which is really stupid; it's chasing moving goal posts.”
AI models identify long-hidden software vulnerabilities
“So OpenBSD is an operating system... it's been around for a long time. It's kind of on the front of the Internet for many corporations. It's used in firewalls. So it's facing the hackers all the time. A guy named Niels Provos had written some code in 1998 and he made a mistake. Nobody noticed that mistake for over 27 years until Mythos took a shot at it.”
“AI is to white collar work and the cubicles and the office parks, what the robot arms were to the factory floors. You just don't need as many whippersnappers making your PowerPoint decks and your Excel spreadsheets and learning corporate lingo. When you go into some of these large corporates, you can sense that there are a lot of people whose jobs are not super vital to the operations and growth of that business.”
Older expert pianists match younger experts on skill tests
“So this study, they took older and and younger expert and amateur pianists. So so you have, four groups. Right? Old and young, expert and amateur. And what they found was that when they were looking at complex skills related to piano playing, the older and younger well well, the experts outperformed the amateurs, I as you might expect. But the older and younger experts, and so, like, the the younger experts were in their twenties and thirties, the older experts were in their fifties and sixties. They performed just as well as each other.”
“Many stable coins today will do things A, like controlling the velocity at which something should be minted. So maybe there shouldn't be more than 10 million mints within an hour. The max is 10 million per hour. And if you want to mint 80 million, it takes you a day, which is totally a legitimate choice.”
Nicotine boosts focus acutely but creates tolerance and withdrawal
“So nicotine can acutely boost cognitive function. There are some studies that kinda suggest that chronic nicotine users have a lower risk of of, Parkinson's disease in in particular long term. I will say that we know nicotine is addictive. We know that it it produces tolerance and withdrawal. So you need higher doses to have the same effect, and then you get significant withdrawal if you stop taking it.”
“His answer is, just put more interesting stuff in front of the camera. Make what's in front of the camera more interesting. And the equivalent of that, at least for me as a non-fiction writer, is doing interesting things. Go out in the world, do interesting things, or observe interesting things in real life and write about those things. Do experiments, et cetera. Anything that is analysis-based is relegated to the machines at this point. They're so good. AI, broadly speaking, LLMs being one manifestation of that are just too good. They're so good. So do interesting things and write about them.”
“My my definition of of a healthy brain is, is a brain that does what you wanted to do when you wanted to do it. And it's kind of less, like, slightly adapted from there's, like, a WHO brain health definition or something. But I put it in put it that way because each of us wants our brains to do different things.”
“Even if you are using AWS and you're using their secret manager, you can have two-fac on AWS, you can have biometrics, and of course, on top of everything, if you care about security, you don't want to have a single key that ever has unlimited permission to do an infinite mint. So even if one person in the organization is compromised, you want it to at least require several more.”
“The CEO I had breakfast with who's freaking out, he said three words to summarize the situation: 'Capital displaces labor.' And then he put in parentheses 'with the help of AI.' But he's like, now it's clearer to him than ever because his company is roaring in terms of performance and stock market price, and he's cutting people right and left.”
Government-funded job retraining has largely failed
“When I looked up the studies as to how effective government-funded retraining programs were, the efficacy range I found was 0-15 percent, with 15 percent on the high side. Trying to train workers who've been laid off in various jobs to compete against AI strikes me as a loser. It strikes me as the next generation of 'learn to code' which is really stupid; it's chasing moving goal posts.”
Exploitation speed has collapsed from years to days
“Eight years ago, the average time between a bug being found and a hacker using that bug in a cyber attack was 847 days. Now it's like within a day. It's not rocket science, but it takes time for a human to do it. You have to have a certain level of expertise. AI has absorbed all of that.”
Reductionism fails — you can't fix a brain like a radio
“So in this essay, he's talking about, you know, if if you gave a biologist a radio, what they would do is they would, they would take it apart. They would look at all the individual components. They would count them. They would look at their colors. They would, like, change the color of the transistors to see if that affects their function and all this kind of stuff. But in reality, what they would have is just like a bunch of bits, and they wouldn't necessarily know how they fit together.”
“When you look at, like, brain volumes, you again, you see something similar. So lots of studies now done in the UK Biobank where if you're averaging up to one UK unit of alcohol a day, which is eight grams of ethanol, which is about half of a US standard drink, which is 14 grams of ethanol, If you're averaging that amount, no real effect on brain volume or dementia risk. If you're drinking more than that, brain volume start to decrease and dementia risk starts to increase.”
“It's that you have a certain threshold, let's call it $35,000 at the poverty level. If you make less than that, then we true you up to that level. If you make more than that, then it's as it was. That's the thing that I'd be thrilled with. I'd be thrilled with a higher child tax credit. I'm thrilled with anything that alleviates poverty.”
“AI is to white collar work and the cubicles and the office parks, what the robot arms were to the factory floors. You just don't need as many whippersnappers making your PowerPoint decks and your Excel spreadsheets and learning corporate lingo. When you go into some of these large corporates, you can sense that there are a lot of people whose jobs are not super vital to the operations and growth of that business.”
“My 13-year-old has come to me and said, I think I'm gonna have an AI girlfriend, not a human girlfriend, because it's gonna be a lot easier. Childhood is hard enough if you don't have trillion dollar companies trying to prey on your brain and your soul. If you show me a household where the kids aren't on screens, they have a much, much better shot at flourishing.”
Alphabet holds a unique full-stack advantage in AI
“Alphabet is an interesting position to, in some respects, kind of own the full stack. Engineers aren't going to like that I'm using that term, but they have distribution. They have hardware in terms of TPUs. They have incredible unparalleled access to information. They've got Demis Hasimus and DeepMind internally. They've got the ability to spin things out like Waymo. There's just so much going on within Alphabet that I find it very fun and terrifying to take a close look at. And I say that also because it is completely unclear how exactly Google compensates for or plans for shifting to some type of ad revenue from AI generated responses.”
“My 13-year-old has come to me and said, I think I'm gonna have an AI girlfriend, not a human girlfriend, because it's gonna be a lot easier. Childhood is hard enough if you don't have trillion dollar companies trying to prey on your brain and your soul. If you show me a household where the kids aren't on screens, they have a much, much better shot at flourishing.”
The home environment predicts brain outcomes more than hospital trauma
“But then when we do these big studies looking at what determines, or what predicts how well their brains function later on in childhood, the biggest predictor or the most important factor is the home environment. And there are different ways to to measure this, but it's basically then related to socioeconomic status, reading, education, nutrition, all that kind of stuff that happens in home is, at home is more important than what happens in the hospital. So, like, even if you had an imperfect start to life, the the environment you go back to is the most important thing.”
“Placement rates have collapsed for particular populations, particularly computer science grads coming out of various universities, where you had programs that had placement rates of 94% at high salaries. And those have flipped. Now you have computer science grads out of great programs looking for six months and getting AI-fueled responses. To me, that population is the canary in the coal mine.”
Stimulants make you feel sharper while performing worse
“What I will say is that there are lots of studies that show that when we use stimulants to improve cognitive function, we think we're performing better, but often we're actually performing worse. So we've they've seen this with caffeine. They've seen this with, methylphenidate and many of the others. And it it's the same with creativity and cannabis. Like, you think that your brain is working much better, but then if you, like, objectively score cognition, it's actually worse.”
Proof of reserve oracles mitigate stablecoin de-pegs
“Also, there are DeFi Native solutions to do this. So there are things like proof of reserve oracles. So you have a separate oracle that is basically tasked with understanding the value of the reserves at any given time. And there's just an on-chain check. There shouldn't be a gap.”
AI models identify long-hidden software vulnerabilities
“So OpenBSD is an operating system... it's been around for a long time. It's kind of on the front of the Internet for many corporations. It's used in firewalls. So it's facing the hackers all the time. A guy named Niels Provos had written some code in 1998 and he made a mistake. Nobody noticed that mistake for over 27 years until Mythos took a shot at it.”
Lactate is the key messenger from intense exercise to BDNF
“However, lactate gets straight into the brain through monocarboxylate transporters, the same transporters that take ketones into the brain and switches on the production of BDNF. So lactate seems to be this key, like, messenger from, like, more intense exercise through to the production of stuff like BDNF in the brain.”
Identify psychedelic red flags through adverse event knowledge
“Specific to clinicians or practitioners, ask them what types of adverse events they've seen. What are the most concerning adverse events that they've seen? A simpler way to put that is, how do you handle freak outs? What do you do when somebody really loses their shit? And if their answer is, people don't lose their shit, there aren't any adverse events, they're either lying, delusional, or very inexperienced. Maybe all three. Those are not mutually exclusive. So I find that to be a pretty quick, necessary but not sufficient way to use a particular line of questioning to separate seasoned practitioners who are honest from those who are neither of those things.”
AI models identify long-hidden software vulnerabilities
“So OpenBSD is an operating system... it's been around for a long time. It's kind of on the front of the Internet for many corporations. It's used in firewalls. So it's facing the hackers all the time. A guy named Niels Provos had written some code in 1998 and he made a mistake. Nobody noticed that mistake for over 27 years until Mythos took a shot at it.”
