NEW🚨 Inside the Company That Evaluates AI Labs; Who Will Reach Recursive Self-Improvement First? @JTLonsdale & @vvkgopalan sit down with @RayanKrishnan, founder & CEO of @ValsAI They cover: - How Rayan used Fable to solve a 370-year-old cipher - The science behind model evaluation / can you stay ahead of the labs? - Should labs self-regulate? What's the role for independent evaluators? - Which lab will reach RSI first — and what happens next? - How do models try to deceive evaluators? Episode 167 out now! (00:00) Episode intro (02:00) Palantir, Machiavelli & lessons from Alex Karp (07:50) Solving a 370-year-old cipher / Is AI making new discoveries? (12:10) How do you measure AI abilities across finance, legal, etc? (15:30) What is the science behind model evaluation? (18:00) The importance of independent evaluation (23:35) Evaluating Chinese models & addressing distillation (26:45) Will Anthropic reach Recursive Self-Improvement first? What happens next? (28:30) How do models try to deceive evaluators (33:00) Why Rayan remains optimistic about AI

NEW 🚨 The Tech Leader Who Is Shrinking Government, Exposing Billions in Waste & Bringing Back Merit @JTLonsdale sits down with a16z Managing Partner-turned-OPM Director @skupor at @uaustinorg - Why leave tech to be attacked in DC? - How did he reduce the size of government for the first time in generations? - How is he bringing merit back to the civil service? - What is he doing to incentivize our best minds to serve in government? - Who should join the U.S. Tech Force? Episode 165 out now! (00:00) Episode intro (01:40) a16z's first employee; why leave tech to be attacked in DC? (06:27) Biggest federal workforce reduction in modern history (08:45) How the left killed merit-based hiring (14:40) How Scott is bringing merit back to DC (15:50) Lessons from DOGE (18:20) Hidden costs of outdated systems; Billions for contractors (23:40) Building a US Tech Force and hiring for skills (26:20) Eliminating degree requirement to get the best talent (27:16) Why should a young talented person serve in government? (30:28) Optimism for DC; rejecting the blackpill

NEW: @JTLonsdale on @RuthlessPodcast says AI labs should be held to a higher standard of liability: "I disagree here: Andrew Ferguson runs the FTC, and I think he's great, a very smart guy. But he said something last week like, "It's just like a gun. If you buy a gun and shoot someone, you have all the liability." I don't fully agree with that for models, because it's not quite the same."

NEW 🚨"Our generation has it so much easier in what you can build and create [with AI], if you're willing to work... There are 2 million small businesses that Baby Boomers have to sell right now..." Politicians aren't supposed to say that. You're supposed to say, 'You've been mistreated, you're a victim.' But it is an awesome time to build if you're willing to work... @JTLonsdale on the @RuthlessPodcast with @ComfortablySmug, @HolmesJosh & @MichaelDuncan Josh Holmes: Do you feel like this is a new American economic revolution? Joe Lonsdale: You can't help but be optimistic when you see the capabilities of these models and what's going to happen with productivity. I'm a fan of Kevin Warsh. I've known him a long time. He's the chairman of the Federal Reserve, and one of the things he'll be doing in the next year is using all the staff there to measure productivity in our economy, because that's a huge part of what's happening here. The only way we all get wealthier is if productivity goes up, and it's very clearly going to go up a lot. I'd be totally shocked if it doesn't over the next few years. When it does, that's going to create a lot of wealth and a lot of jobs, if we can manage not to rip each other apart with our crazy populists and keep this an actual capitalist economy where you're allowed to build things and innovate. There are 2 million small businesses that baby boomers have to sell right now. It's so much more fun to run a small business using even a little bit of AI. Our generation has it so much easier in what you can build and create, if you're willing to work. It's not popular to say. Politicians aren't supposed to say that. You're supposed to say, "Oh, you're so tough, you've been mistreated, you're a victim." But it is an awesome time to build if you're willing to work.

NEW 🚨 Inside Anthropic — AGI, AI Optimism, Regulatory Capture, Open Source & More!  @JTLonsdale sits down with Anthropic's @_sholtodouglas and @the_marwell - Have we reached AGI — and what does it mean? - What are the most exciting breakthroughs at Anthropic right now? - What are the biggest risks and concerns? - What is their stance on open source? - Is Anthropic angling for regulatory capture? - Do AI researchers share same values as the heartland?  Episode 164 out now!  (00:00) Episode intro (02:00) Self-taught ML; Unlikely paths to Anthropic (04:30) What are the most exciting things you're working on? (08:04) Have we reached AGI — and what comes next?  (10:10) How to measure AI progress (14:20) Is biology the next great frontier? (16:50) Career advice & the decade of the generalists (19:30) What could go wrong? What are the biggest AI risks? (23:11) Offense vs defense; how technology evolves (25:08) Answering the regulatory capture critique (30:45) Anthropic's stance on open source (31:50) Would you slow down and let China go ahead? (33:05) Anthropic's view on distillation (40:12) Is a post-scarcity world achievable? (48:54) How do we preserve values, tradition & community? (55:00) Predictions for 2028

🚨NEW: @JTLonsdale & @SquawkCNBC clash on AI and wealth inequality. Joe Lonsdale: 2030s is the disinflation decade... we're all going to be really wealthy if we don't break this. Joe Kernen: Everybody's going to get wealthier? Because only the wealthy have been getting wealthier... Joe Lonsdale: That's not true. Let's go back: 1870 to 1900, the Second Industrial Revolution, the last time productivity shot up a lot, the average American, the average working class American, was twice as wealthy in real terms within one generation... Andrew Ross Sorkin: Tell me where you land on "We've got to slow everything down, because if we don't, we could end up dying... Joe Kernen: or worse... Joe Lonsdale: or worse, the companies go bankrupt  Andrew: versus we've got to go full steam ahead. No matter what. No matter what the case. Joe Lonsdale: I love the meme where it's Dario and Sam saying, please regulate us. And Trump says, keep building, nerds.  These companies have a lot of liabilities. These companies need to slow down if they see themselves causing damage. I'm not worried about any existential risks in the near term. I 100% could see these companies accidentally causing some problems with hacking with cyber, for example. They have a responsibility to slow down themselves, and we should be holding them liable if they cause any damage. And so it makes sense that they would be very careful about that. But we were just talking... 2030s is disinflation decade. We're going to all be really wealthy if we don't break this. So don't slow down too much is my view. Joe Kernen: Go back to this again. You told me that 2030s are going to be the decade of disinflation. We're going to all get rich, and we're not going to need to raise taxes, and we're going to start growing faster. The deficit will come under control. This is going to happen? Joe Lonsdale: The AI is really good for rooting out fraud and grift in the government. It's really going to bring down health care costs dramatically. It's amazing for bringing back manufacturing. We have companies already bringing back lots of manufacturing here. That's only about to accelerate. We're at what's called a GPT-3 moment right now for robotics, right. World models, which are a new type of AI -- how do things move and interact in the real world in general terms... Joe Kernen: So this will help one of our most daunting fears of the $40 trillion and the out of control deficits we're running now, this will come under control from AI?  Joe Lonsdale: If you get productivity up to 3, 4 or 5%, what you 100% should be able to do with these things, if not significantly higher, that solves the deficit problem. This is an amazing thing for a civilization; this is the answer is to get growth higher, and it's how everyone gets wealthier, it's doing more with less. It's going to do it. Joe Kernen: Everybody's going to get wealthier? Because only the wealthy have been getting wealthier... Joe Lonsdale: That's not true. Let's go back: 1870 to1900, the Second Industrial Revolution, the last time productivity shot up a lot, the average American, the average working class American, was twice as wealthy in real terms within one generation. And that's what happens when productivity goes up. Everyone gets wealthier. We we teach economics very poorly in this country.

NYT attack on @JTLonsdale: he's an outsider, tough on violent criminals, defends capitalism, rails against the status quo, and isn't politically correct. Also NYT: somehow he's the most effective change agent in homelessness today.

EPISODE 161: The Robotics GPT-3 Moment @JTLonsdale sits down with @peteflorence Pete and @GeneralistAI are changing everything we know about robotics: -- robots trained on new tasks in seconds (seconds!) -- self-taught ambidexterity & other breakthroughs -- 10X more science and how the world will change (00:00) Episode intro (01:30) Bay Area to MIT and DeepMind (03:55) Pete's epiphany: robots sitting still (08:30) Why intelligence, not hardware, is the next frontier (13:20) The GPT-3 era of robotics (14:55) Why leave DeepMind? (16:50) How to train robot brains (18:30) Emergent ambidexterity (20:40) LLMs vs robot world models (23:50) GEN-0, GEN-1 & scaling laws in robotics (27:28) 10x more science per year (32:50)  Skilled trades, reindustrialization & future of robots

"I remember getting there on day one and reading the [SpaceX] employee handbook. On the top of page one was just a single line in bold that said: this is not a science experiment." @ScottNolan was an early employee at SpaceX He explains what made the culture and mission unique.