USE AI

“With LLMs, the barrier to entry into an area of programming that you've never done before is nil. I was annoyed at some remote control functionality that I have in my home AV system. So I decided I'm just going to build an iOS app to completely get rid of my stupid remote and have a remote on my phone. Normally, a year ago, I have to learn how to build iOS apps, I have to learn how to interface with this audio video system that I have. This time around, I was like, well, I can just ask my favorite agents to go research stuff and set it up for me.”

“At the end of the day, we're balancing resources like compute power, data, and human effort, and we need to be able to balance all of those. We thought a lot in the AI space about compute power. We haven't thought that much about the human effort one. And I think with reinforcement learning, we minimize that. We allow it to explore and do its learning without the kind of human labeling.”

“Computer science is a science... It's also an art form. Like if you read code, or maybe we used to read code for fun, some code is more elegant than other, and you would readily see it. That's a beautifully written piece of software. I'm like, that's kind of garbage, but it does the job. So there's a matter of taste and elegance in programming, and it's also a craft. When you actually look at well-made production-grade code, you can identify if you know what you're doing, like that's a well-crafted piece of software.”

“What they found is that if one evokes the release of adrenaline through this arm into ice water approach, the information that they read previously, just a few minutes before, was remembered, it was retained as well as emotionally intense information. This had to be the effect of adrenaline released into the brain and body, because if they blocked the release or the function of adrenaline in the brain and or body, they could block this effect.”

“We are actually coming out, we're going to be more, we actually have some incredible diagnostics and predictors for actual diseases and we're moving down the reimbursement pathway. We have a predictor for dementia, of are you going to get dementia in the next five years, that has an AUC of over 0.9—so over 90% ability to discriminate that.”

“Each employee in our company is required to be highly competent in the use of AI. There is a minimum of one AI training per week that 100 percent of employees must attend, and you have to provide proof that you use AI for one hour a day. We have currently over 400 documented helpers or tools that we have built for AI, and it allows us to do things that human beings simply cannot do by themselves.”

“Typically, what you would do is you would aggravate the scalp. This would be using something like micro-needling. Dermarolling would be another option. And then you apply and rub in GHK copper peptide. It's not hard to get your hands on GHK copper peptide... it is very small, I think it's about 50 daltons in size. So as a peptide, it's small enough to be absorbed transdermally.”

“I became obsessed in 2008 and 2009 with the idea of ranking cities for real estate investments. How about bringing that level of clarity, that level of objectivity to ranking cities for real estate, which cities are likely to be more profitable? You can never say it with any level of certainty, but you can certainly improve your chances pretty substantially of making money if you are data driven.”

“It was literally the same day I had received this proposal from this dev shop in Ukraine for $105,000. They had actually sent over, we'd already done some architecture discussions and they'd sent over a PDF with a proposal, but here's how we're going to build the whole thing. And so I was like, I wonder if I could just drop this PDF into Replit. So I just drag it in and said, build this. And 10 minutes later, I had, you know, at least the first version, that preview.”

“To the average user, you sort of know that you're interacting with OpenAI, but you don't actually realize how much is going through their servers and their systems. Behind the scenes, there's enormous kind of layers of technology, and, actually, that isn't necessarily the best way to build things as maybe listeners of this will already know. We allowed all of the power to accrue to that small number of companies who built that infrastructure.”

“Niantic isn't just a gaming company; it's a 3D mapping company using the world's players as its data collection army.”

“Now, memory is simply a bias in which perceptions will be replayed again in the future. Now, this might seem immensely simple, but it raises this really interesting question, which we talked about before, which is why do we remember certain things and not others? Because according to what I've just said, as you go through life, you're experiencing things all the time.”

“Your DNA makes up about 20% of your health outcome, but the other 80% is what you do, what you eat, how you sleep, your stress levels. With epigenetics, you're measuring that. That other 80%, you're actually looking at how much every gene is turned up and turned down. What we can predict with epigenetics is limited by our imagination and the size of the data set.”

“It is the presence of high adrenaline, high amounts of norepinephrine and epinephrine that allows a memory to be stamped down quickly and far and away different than the idea that we remember things because they are important to us or because they evoke emotion. That's true. But the real reason, the neurochemical reason, the mechanism behind all that is neurochemicals have the ability to strengthen neural connections by making them active just once.”

“AI users have 9.4 times higher code churn than non-AI users, and Pharaoh AI found code churn increased 861% under high AI adoption. The productivity gains from AI coding are real, but they're also a fraction of what the output numbers suggest, right? If you're like, I can write a million lines of code a day now... well, a big chunk of it has to be rewritten or fixed.”

“You can describe what you want, a pitch deck, a one-pager, a landing page prototype, and Cloud generates a first draft. It can also read your company's code base and design files to apply a consistent design system across all of your outputs, which is actually, I think, the more interesting piece.”

“GHK copper peptide, abbreviated GHKCU... works through three mechanisms. It stimulates what are called fibroblasts, which is the growth of new cells. It promotes new blood vessel formation so hair follicles receive necessary nutrients. And then it inhibits something called TGF beta to prevent follicles from prematurely shrinking. It also supports what are called dermal papilla cells to stimulate hair growth.”

“2023 also saw the roots of a global banking crisis arise out of four American regional banks, the two largest being Silicon Valley Bank and First Republic Bank. 2021's inflation surge moderated in 2023, while the Federal Reserve continued to raise its interest rates in the first half of the year.”

“Additionally, the latter half of the year saw many large mergers and acquisitions, some of the largest announcements being in oil and gas with ExxonMobil's purchase of Pioneer Natural Resources for nearly $60 billion and Chevron's acquisition of Hess Corporation for $50 billion, both in October and pending regulatory approval prior to closure.”

“What we found was there was strong correlation between job growth, income growth, home price growth, and crime reduction, and population growth. These five areas seem highly correlated to real estate profits. All of them make sense initially, but the devil is in the details, because the question is which one's better and how much better? How do you quantify something like this?”

“If we could actually use those devices for some of the tasks that the company would do, like training machine learning models rather than just using them, you could have far more kind of efficiency gains over the world's resources if you built things in that way. So from a purely technological perspective, there's actually a better way of using the world's resources than the current kind of centralized approaches.”

“Socialism sucks. It is terrible. The only people who do well in redistribution of wealth are the ones doing redistribution. It's fundamentally corrupt, and there's not enough bad things I can say about socialism. So I think capitalism works exceedingly well, especially when the competition is encouraged and allowed to flourish... We've had thousands of years to evolve a fairly good economic model, and US is most certainly 1A in the right way to do it.”

“The thing that's happened with the LLM powered models is you can now go from, I have an inkling in my head to, I'm going to build a prototype to, I'm going to, not necessarily always on your own, but with a bunch of co-conspirators to build something that's production ready in record short time. And if you do know what you're doing as somebody with a sense for how to build software well, you don't need to vibe code yourself to a prototype and then hand it off to somebody who knows what they're doing. You can actually build something amazing and just ship it.”

“The attacker actually broke into Context AI, which is an AI tool that Vercel, one of their employees was using. And from there, they got into that employee's Google workspace and from there into Vercel's environment. So there's kind of like this multi-step approach that they took in order to get in.”

“We found 1600 biomarkers we could predict when we measured them against their traditional versions. About two-thirds performed better. So better hazard ratio or odds ratio. So we found about a thousand, quote-unquote, upgraded biomarkers. And that's how we developed the TruHealth test. It's the top 100 or so of the ones that actually are better at predicting disease than the traditional LabCorp request versions.”

“We reach a point where the machines are autonomous to be able to make their own decisions. They have their own on chain identities. They have the ability to update their own objective functions and reward models, which in combination with crypto economic property rights basically makes them sovereign economic actors. All this will be coming in the next twelve months.”

“Pi 0.7 can perform tasks it was never specifically trained on by composing skills it learned in other contexts. In some broader testing, the generalist model actually matched specialized models on jobs like making coffee, folding laundry, and assembling boxes. Researchers said that the generalization ability was really surprising to them.”

“I think the net IQ of the world is about to go up, like, 50 points... The average IQ with AI in your ear at all times is about to go up to 150, which is, like, north of the genius definition. So very, very soon, the willingness to put up with random obnoxious things, like low quality or crappy terms or products that pretend to be something they're not, like, all of that is just getting flushed out very quickly. And companies that have business models that are buried in a fine print of some kind are all in for a very rude awakening.”

“We're sitting at about 4.5 million in annual recurring. I started this business with just $400. As a one man band, I was able to take it by myself with all of my AI tools to a million bucks. Let me just tell you exactly how to do this, right? And everyone's looking for, well, what do I do? How do I do it? Here's my process.”

“Crypto allows us to do that entirely programmatically. And so that's very efficient. If it's just humans, it's great. It makes it more efficient. If it's machines, it's essential. Because a machine can't go through the court system. It can't operate through those human world trust mechanisms. It needs programmatic trust mechanisms. And so crypto becomes essential if you start swapping in machines for humans in those interactions.”

“It depends like where you sit in the stack, right? Like um, now for Meta, it's like all it's all like creatives. Um, a lot of people are going to cycle into like AI creatives and then that's going to burn out. And then it'll be the next thing. Might be UGC creative uh, creatives again. Maybe affiliates. But it really like any good marketer uh, is going to figure out what the next arb is and they're going to maximize it. So, if you're like a real marketer, not just like one shot one shot wonder to understand Meta, then yeah, you're going to be able to you're going to be able to sell and you're going to be able to sell a lot faster in this new era.”

“I think marketers are primed to win the biggest in this space. Cuz developers are obviously clapped. Um, to give you guys some insight, we recently stumbled upon a tech walk in Vancouver. And half these guys are looking for a job. The other half is saying they like to think with their brain um, rather than use AI tokens. And I think I think Tony said something like if you're thinking, you're not using tokens, which is like NGMI, like automatically, right? Um, Yeah, I mean, there's a there's a thinking machine for us. Like why use that type of like deep thought, deep thinking?”

“Previously, they had Image 1.5. Basically, it's really good at doing something that the last model, 1.5, was pretty bad at, which is rendering text. If you ask it for a poster, you know, it would always get this kind of like elfish looking, I don't know, the letters weren't that great. Anyways, 2 in their words is a step change.”

“Just like an intense exercise, you're going to have liver function tests from acute stress after and CRP, all these things, but we know then two days later, it was a good thing for us. It really looked like a picture of just like there was a little bit of designed overreaching or some other stressful thing during that one test.”

“I'm actually personally bullish right now on um, like one-person agencies. Like if if if I had to make my first freaking 10k a month, I'd just be one guy with AI, armed with AI, probably like a Claude Max. And then just try to figure it out from there. Um, and I'd post every day on socials, like one shorts one short a day, and just go that that route. It's probably cuz I know that route. Like I know that method pretty well, so I feel like I'm I could run it up. Like I don't know, gun to my head type of thing.”

“As I was talking myself out of sending this email to this person, I saw this Ted talk and the talk was like, 'tomorrow never comes.' Everyone says they're gonna go for the bullseye and they just never do. And so I was like, let me just send this email, and Ryan responded.”

“I got an email and I thought it was a total joke because the subject was like 'Starbucks has requested your song.' And then I researched the person who sent it and he was a sync placement dude based in New York. It was really validating for me as coming out as a lyricist.”

“What's happening is that you're encouraging the firing of particular chains of neurons that reside in a particular circuit, right? So a particular sequence of neurons playing neuron A, B, C, D, played in that particular sequence over and over and over again. And with more repetitions, you get more strengthening of those nerve connections.”

“A lot of the money is moving out of software into fiscal goods. So, I'm like pretty bullish on physical products right now. Unlike software, there's like software really has no moat right now. I've seen Jackie Vibecode like absolute trash. Um, Yo, chill. If and if Jackie can do it and if I can do it, uh, there's going to be like a like 10 million other degenerates that are going to code up the same [ __ ] So, I think like at least for venture and like private equity and growth equity firms to invest in software is going to be very challenging moving forward. It's an unknown space and money's going to money's going to rotate to the fiscal.”

“Niantic isn't just a gaming company; it's a 3D mapping company using the world's players as its data collection army.”

“We reach a point where the machines are autonomous to be able to make their own decisions. They have their own on chain identities. They have the ability to update their own objective functions and reward models, which in combination with crypto economic property rights basically makes them sovereign economic actors. All this will be coming in the next twelve months.”

“At the end of the day, we're balancing resources like compute power, data, and human effort, and we need to be able to balance all of those. We thought a lot in the AI space about compute power. We haven't thought that much about the human effort one. And I think with reinforcement learning, we minimize that. We allow it to explore and do its learning without the kind of human labeling.”

“Like you can literally vibe code a solution. You might have a task that was going to take you six hours, but instead you could take 30 minutes and vibe code a tool that will do that task for you. And you may never use that piece of software again, right? But you wouldn't think like that back in 2020, right? Or before vibe coding was a thing, right? Back when I started my company in 2016, that was not an option for me.”

“So we've turned the corner of saying, okay, yes, this STE model still is there, but now it is assisted by software. Now we really believe that GenAIPI becomes the AI infrastructure for our clients. And that becomes also much stickier. No one wants to turn that off. It takes time to get it integrated into the systems and to work with NetSuite versus QuickBooks versus this system and HubSpot versus Salesforce, and you have to do some connections and things like that.”

“They're recovering from the trauma that they experienced living in captivity. And for them to open up and trust you while you are there with them, helping them work through it, it's indescribable.”

“We found 1600 biomarkers we could predict when we measured them against their traditional versions. About two-thirds performed better. So better hazard ratio or odds ratio. So we found about a thousand, quote-unquote, upgraded biomarkers. And that's how we developed the TruHealth test. It's the top 100 or so of the ones that actually are better at predicting disease than the traditional LabCorp request versions.”

“We need more than 1% annualized population growth for rents to increase aggressively. 3% job growth in a certain market can lead to very aggressive rent growths; 4% is phenomenal and 5% is rarely achieved, but it's absolutely incredible because if a city has 5% job growth, everyone at that point is employed and everyone is very aggressively looking to rent or buy.”

“The number one intervention over everything I'll ever tell you about—whether we studied this supplement or this medication—none of that is close to the strength of your social relationships. That study was so definitive that anyone who's biohacking and not focusing on that is wasting their time with everything else.”

“Factory... just closed $150 million Series A at a $1.5 billion valuation. Morgan Stanley isn't going to let some random developer tool run inside their network unless it's built with their compliance and security posture in mind. I think that's basically the gap that factory is filling.”

“They have in the past several weeks, actually, fresh leading into this podcast, developed an aesthetic line, meaning they've taken Mew cells, which are specialized regenerative cells that are well recognized for their inherent renewal capacity in exosomes, which serve as extracellular messengers to support better cell communication. You can use these cells in combination with exosomes for halting hair regression and actually encouraging the growth or the generation of new follicles.”

“Just like an intense exercise, you're going to have liver function tests from acute stress after and CRP, all these things, but we know then two days later, it was a good thing for us. It really looked like a picture of just like there was a little bit of designed overreaching or some other stressful thing during that one test.”

“When I first started, there was so much tenacity in me that I would say to any new artists: run that out first. Do everything you can. You have to use that life force in you because you learn so much about yourself and what you want and what you are.”

“Modern FUE, which is called Follicular Unit Extraction, is the gold standard for a re-seeding hairline. That involves taking individual hair follicles from a donor area... and the result can be pretty undetectable with Follicular Unit Extraction. It is only moving existing follicles from an area of your body. It can be the back of your head, for example, to the thinning areas.”

“To the average user, you sort of know that you're interacting with OpenAI, but you don't actually realize how much is going through their servers and their systems. Behind the scenes, there's enormous kind of layers of technology, and, actually, that isn't necessarily the best way to build things as maybe listeners of this will already know. We allowed all of the power to accrue to that small number of companies who built that infrastructure.”

“You could use like a like a drop shipping model to test out an idea first. So, I think the best way to test an idea to see if it works and to see if you have the right unit economics, it's like yeah, drop ship it first, run some ads to it, see what the customer acquisition cost is going to look like. Um, and scale it up. You could you could basically launch a site within like like under a week. You probably depends on how cracked you are. With AI for sure. Like website, you can AI with it. You can get a bunch of like fast creatives, static using AI or or UGC.”

“It does not stop the progressive loss of your native hair. Meaning that if you have a lifestyle that's full of, let's say, inflammation, poor blood flow to the scalp, poor genetics... all of that will continue to occur, even if you get a hair transplantation. If you don't address the underlying cause of hair loss after a transplant, you'll continue losing the hair that wasn't transplanted and eventually look pretty patchy again.”

“In the short term, the market is a voting machine, in the long term, it's a weighing machine. And I'm a big fan of reminding myself, if the price is going up a lot, you can overlook that if price is going down a lot. It's important to remember, but either way, you're working for the long term. If you're willing to drop everything and look at the short term, you're joining the group of investors that I said I don't care about. I certainly don't want to do that.”

“He said that AI is now generating more than 65% of new code at Snap and the company expects to cut over $500 million in annualized expenses by the back half of this year. So they are cutting in a massive way. The stock jumped almost 8% on the news, and I think that tells you exactly how Wall Street is feeling about this tradeoff.”

“We are actually coming out, we're going to be more, we actually have some incredible diagnostics and predictors for actual diseases and we're moving down the reimbursement pathway. We have a predictor for dementia, of are you going to get dementia in the next five years, that has an AUC of over 0.9—so over 90% ability to discriminate that.”

“Conditioned place preference, as with condition place avoidance, depends on the release of adrenaline. It's not just about stress. It's about a heightened emotional state in the brain and body. You can get one trial learning for positive events, condition place preference, and you can get one trial learning for negative events. This turns out all to be true for humans as well.”

“I really just looked at what would they have to pay if they were bringing someone internal, right? You know, I believe that every company, every single company should have somebody in charge of AI. Somebody needs to be in charge of AI and their job is to go around and keep people accountable. In order for me to actually provide ongoing support, and really this is what drove it in the beginning. In order for me to actually provide ongoing support, it's got to be worth my time. So I'm going to charge it at a good employee, 15K a month to start.”

“Your DNA makes up about 20% of your health outcome, but the other 80% is what you do, what you eat, how you sleep, your stress levels. With epigenetics, you're measuring that. That other 80%, you're actually looking at how much every gene is turned up and turned down. What we can predict with epigenetics is limited by our imagination and the size of the data set.”

“Crypto allows us to do that entirely programmatically. And so that's very efficient. If it's just humans, it's great. It makes it more efficient. If it's machines, it's essential. Because a machine can't go through the court system. It can't operate through those human world trust mechanisms. It needs programmatic trust mechanisms. And so crypto becomes essential if you start swapping in machines for humans in those interactions.”

“OpenAI just massively beefed up Codex for desktop control, memory, and in-app browser, and over 100 plugin integrations. Basically, this is them swinging directly at Anthropix Claude Code and Claude Cowork, and I think it matters a lot for where editing agents is going to be going in the future.”

“Because I had built this image of I am the General AI Proficiency Institute, and not only that, I'm a.org, right? And there's just this added level of trust to a.org and an institute and whatever, and it enabled me to get the big deals with businesses. They said, oh, well you obviously are an institute. Yeah, we'd love to work with you. I mean, I got those kinds of things. I had governments from Australia reach out to me and say, hey, seems like you're the worldwide standard on this.”

“Everything even still now is still lyric driven. So I just create music that sets a tone or an emotion for the emotion I'm feeling in the poetry. I rarely have come up with a musical piece first, and I am constantly driven by the lyric because I have tons of ideas.”

“I started off with no musical background, like at all, like never touched an instrument, never really thought I would ever in my whole life. And then in my last year of college, I was a literature major, and I was like, I'm gonna be a famous poet. And then I met this old soul singer named Malcolm Hayes who lived above Harvard and Stone.”

“I think companies that have built software orally and just sell that software are very vulnerable. The bar for quality of software is going up rapidly. The, yeah, it kind of sucks and has bad interface, but it really serves an important function and I can't be bothered to hire engineers or build the same thing myself. That excuse is gone. If you really hate some piece of software that you're using and it's just software, it doesn't have some deep proprietary data, proprietary source of value, it will get replaced.”

“The Dunedin PACE. That is my favorite of the biologic age test. It’s developed by Duke and it has actually been shown to be the best predictor of morbidity and mortality period. There was a really big trial called the BASE-2 trial that just came out in the last month. And it is better than Grip Strength, VO2 Max, and all the other things.”

“I basically built an AI coach for myself that I would feed in all my TruHealth data, all the stuff that we're going to talk about, all my training data, my lactate threshold stuff. It would come up with a plan, come up with a supplement plan, come up with everything, and every day, I would upload my workout that I did and it would adjust.”

“The Dunedin PACE. That is my favorite of the biologic age test. It’s developed by Duke and it has actually been shown to be the best predictor of morbidity and mortality period. There was a really big trial called the BASE-2 trial that just came out in the last month. And it is better than Grip Strength, VO2 Max, and all the other things.”

“If we could actually use those devices for some of the tasks that the company would do, like training machine learning models rather than just using them, you could have far more kind of efficiency gains over the world's resources if you built things in that way. So from a purely technological perspective, there's actually a better way of using the world's resources than the current kind of centralized approaches.”

“I basically built an AI coach for myself that I would feed in all my TruHealth data, all the stuff that we're going to talk about, all my training data, my lactate threshold stuff. It would come up with a plan, come up with a supplement plan, come up with everything, and every day, I would upload my workout that I did and it would adjust.”

“I eventually found out, okay, companies really need not just some training and education and show them what to do. They need somebody that is embedded into their company, that is following up regularly and becoming part of that company. And so I built this fractional Chief AI officer model. And the cool thing about that, and that I knew as an entrepreneur was super valuable, was this recurring revenue element.”

“The number one intervention over everything I'll ever tell you about—whether we studied this supplement or this medication—none of that is close to the strength of your social relationships. That study was so definitive that anyone who's biohacking and not focusing on that is wasting their time with everything else.”

“Affirm was founded in many ways to fight all of that and destroy the ridiculous and the exploitive... We don't compound interest specifically to make sure that we can pre-price every loan and say you're borrowing $500 and your total interest chargers will be $25 or $0 if somebody else or merchant is paying your interest for you. And for a long time, consumers were sort of like, yeah, you say that, but there's got to be something in a fine print. One of our core values is no fine print to remind ourselves and the rest of the world that we're not hiding our business model in there.”

“The rise of artificial intelligence and large language models dominated not only the economy but has also been at the root of a Hollywood double strike conducted by Writers Guild of America and a SAG APTRA strike. These were part of a larger phenomenon of labor strikes across the country, in which such large diverse groups, such as Teamsters and Auto Workers won new contracts.”

“As part of the settlement, Clarify has to delete the 3 million photos and any models trained on them. I think that this is something that is absolutely insane. For years, AI companies have been operating kind of in this gray area that, you know, they're like, well, if it was on the internet, we can use it. And the FTC just said, no, you actually can't.”

“It’s strategy, transformation and education. I was already doing that education piece a bunch, right? I had been teaching people and coming in and helping them get up to speed. But I also had been doing strategy as well with a lot of these CEOs before a call or before an education seminar thing I would do with people. I would sit down with the CEO and having, I had the benefit of again, having been a CEO for a decade and so I had experience, I had ideas, I could actually consult at a decently high level.”

“Apple CEO Tim Cook is officially stepping down on September 1st, and John Tarnas is gonna become the CEO. This is really interesting. There's a lot of AI things at play for Apple, and they really haven't delivered on their AI promise.”

“The bottom line is that we end up significantly setting ourselves up for better returns or results when we use data science. There's so much room for analysis, data analysis, for comparisons of various cities and how they're doing in terms of job growth, population growth, income growth, home price growth. You don't always get things right, but we end up significantly setting ourselves up for better returns beyond what typical Excel spreadsheet analysis can do.”

“The dominant political story of the year has been the 270-day-long speakership of Representative Kevin McCarthy, whose slim majority in the House of Representatives has enabled a far-right rebellion to exert more weight over the lower chamber. The battle between the rebellious Freedom Caucus and McCarthy has been at the heart of an averted debt ceiling crisis and the annual budget debate nearly devolving into a government shutdown.”