Tech Brew Ride Home - Is Everybody Racing RSI?

Episode Date: July 29, 2026

Over 1,100 AI staffers signed a letter asking Washington to help pace frontier development. Wall Street got nervous as AI capex ballooned, Silicon Valley's backlash against Anthropic built over open weights, and the argument that this is all about recursive self-improvement either being right on the horizon, or a dead end. AI's finally expensive enough to make Wall Street nervous (The Verge) Over 1,100 staffers from AI companies, including John Schulman and OpenAI's Jakub Pachocki, sign a letter requesting the US government to "pace" AI development (Bloomberg) Signatories to the "Pacing the Frontier" statement include OpenAI's Mark Chen and Wojciech Zaremba and Anthropic's Jack Clark, Chris Olah, and Jared Kaplan, who say the world may need the option to buy time (The Verge) Anthropic faces backlash from Silicon Valley partners, founders, and researchers for competitive tactics, guardrails, and lack of support for open-weight models (WSJ) The Actual Reason Why Google "Fell Out" of the AI Race Changes Everything (The Algorithmic Bridge) Subscribe to the ad-free feed. Learn more about your ad choices. Visit megaphone.fm/adchoices

The first ever all-electric 2026 Subaru Trail Seeker is the EV for the Trail Obsessed, with up to 444 kilometers on a full charge and DC fast charging from 10 to 80% in about 30 minutes in ideal conditions. Plus, ample ground clearance and symmetrical all-wheel drive make the 2026 Trail Seeker the most capable EV Subaru has ever built. Test drive it at your local Subaru dealer or visit Subaru.ca. Welcome to the TechBrew right home for Wednesday, July 29th, 2026. I'm Brian McCullough today. Over 1,100 AI staffers signed an open letter asking Washington to slow down AI development. Wall Street got nervous as AI Cappex ballooned, Silicon Valley's backlash against Anthropic built over open weights, and the argument that this is all about recursive self-improvement, either being right on the horizon or a dead end.

Here's what you miss today in the world of tech. Today's going to be an odd day because I want to sort of devote the whole episode to contextualizing the state of the AI race at this moment through a few different angles. First, the Verge has a pretty good take on the wild swings. The stock market has been making around AI of late, quoting Elizabeth Lapato. Meta, Amazon, and Microsoft will all report their earnings this week, and there are plenty of people who think that, like Google, they will also announce they are spending more than expected on their data center buildouts.

There are a few other things happening at the same time that suggest investors are getting nervous, first of all, people seem to have finally noticed that SpaceX sucks. As of this writing, its shares are worth almost half as much as they were during its peak. Second, investors are nervous about Oracle's data center build-out debt, and it's worth keep in mind that Oracle is the public markets stand-in for Open AI. Third, Nvidia has been encouraging rounds of deal talks worth a combined three-quarters of a trillion $1,000. Even more so than Open AI, is at the center of the circular financing in the AI ecosystem. If it is pumping more money into supporting the AI buildout, that may be an indication that the actual

demand is weaker than expected. Specifically, Nvidia, guaranteeing OpenAI's debt, a deal worth $250 billion is, quote, as much a reminder of funding strain in the AI buildout as it is of demand signal. Billy Long, Global X management's tech sector investment strategist told Blue Bloomberg. On top of all that, a Chinese startup released a new model and people get nervous every time that happens. One reason for that nervousness is that China's biggest constraint is that they, at least theoretically, don't have the same kind of access to GPUs as US companies, and yet their AI systems are still competitive. If that is indeed what is happening, there's an end in sight to Nvidia's and other chipmakers Cash Bananza. What's more,

it may mean that companies are building too many data centers. I've spoken to a lot of smart people who are more optimistic than I am about the AI boom. I have been asking how AI companies plan to actually make money for three years now and I still have not received a satisfactory answer. All of them think we will likely overbuild data centers during this period of exuberance. They also think that a lot of AI companies will die off when the inevitable correction comes. They are invested in the space anyway because they think the companies that survive will make them more money than they will lose on the ones that die off. So the AI boosters are watching for the market top just like everyone else.

they know it's inevitable. It's hard to figure out what the market top is in advance, of course, but some investors are clearly starting to get cold feet about the whole AI thing, and they're moving their money elsewhere, end quote. Next, over 1,100 staffers from various AI companies, including John Schulman and Open AI's Jacob Pachoki, signed a letter requesting that the U.S. government pace AI development. Open AI and Anthropic release statements in support of the so-called pacing the frontier initiative. Anthropics says Dario Amadai and several co-founders have signed the letter, quoting Bloomberg. The petition urges the U.S. to support an international effort that would help set the pace for building more sophisticated AI systems if needed.

The document warns that there is a, quote, real risk that AI progresses faster than people can, quote, understand or control, referencing advances in automating AI research. We request that the U.S. government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development. the petition reads. The petition echoes recent calls from AI leaders, including OpenAI CEO Sam Altman and Google DeepMind CEO Demas Hesabas for a new international watchdog to vet and set standards for cutting edge AI models. Anthropic has also previously suggested the creation of a system in which governments and AI developers could collectively decide when to slow work on AI to stave off risks. It would be good for the world to have the option to slow or temporarily pause AI work that may be dangerous, the company said in June. Those suggestions followed height.

scrutiny of improvements in AI's ability to field more complex tasks with limited human oversight, including spotting and potentially exploiting cybersecurity vulnerabilities and critical software, end quote. And quoting the verge. In the public statement, the employees went on to underscore the importance of some sort of coordinated regulation worldwide so that the ever-intensifying AI race doesn't outpace safety and monitoring tools. To realize AI's potential, industry, government, and society at large may need the option to buy time to address emerging risks, measures and strength and oversight, they wrote. But each company and country is under intense competitive pressure, not to unilaterally slow that acceleration. And today, the world lacks

the technical and governance tools to deliberately pace frontier-wide progress, end quote. Some of the Open-A-I executives or co-founders who have signed the statement include Chief Research Officer Mark Chen, Chief Scientist Jakob Pachaki, co-founder John Schulman, and co-founder Wechek-Zaremba. As for Anthropic co-founder, Jack Clark, co-founder Chris Ola, co-founder Ben Mann, and Chief Science Officer Jared Kaplan, have all signed as well as Boris Cherney, who created and leads Claude, and Ethan Perez, who is Anthropics Alignment Team Lead, opening eyes former chief futurist Josh Akiam, and the former co-lead of its now defunct superalignment team, Jan Liki, also signed. You can see the full statement and a list of

the more than 1100 signatories on a dedicated website titled Pacing the Front Team. here, and quote. Now, more on that context of the acceleration bit of this in a second. But first, I want to come back to that idea that Anthropic is facing backlash from Silicon Valley partners, founders, competitors, and researchers for their alleged competitive tactics, guardrails, and so-called lack of support for open weight models, quoting the journal. At an event shortly after a recent release, Figma chief executive Dylan Field, said Anthropic hadn't been, quote, consistently candid in their communications, according to attendees of the event. The whole industry learned a lot of lessons from how that dynamic played out, said Sarah Sacks,

head of AI at Notion, who attended the event. Some companies might have been overly trusting. Anthropic has also drawn criticism for changes to its data retention policies and for releasing a powerful model that won't answer AI research questions, prompting complaints from some researchers who advocate for more transparency in AI systems. There's also skepticism about the warnings of Anthropic executives about the security risks of so-called open-weight AI, models that are far cheaper and generally allow users to customize and install their own safeguards. Several such models that were produced in China have wowed users in recent months with capabilities that edged closer to frontier AI US systems.

Anthropic makes fantastic products, said Vishal Misra, the Vice Dean of Computing and AI at Columbia University's Engineering School. But the company also makes statements about security and AI risks that don't hold up to scrutiny, he said. They create a scare in the general public around these models, he added, fear from the public gives them leverage to get regulations done in their favor, end quote. A company spokeswoman defended the steps Anthropic has taken as necessary to ensure models are safe and that adversaries can't use them to create severe risks. Anthropic CEO Dario Amadai and other company executives have repeatedly spoken about the risks of

openweight AI systems, which many AI researchers see as far more difficult to control a worrisome possibility if models continue to advance in capability. Some don't have the same security safeguards that leading U.S. models do, AI executives say. Amo Dai urged policymakers to crack down on industrial-scale distillation. Both companies have discussed the challenges posed by Chinese open-weight models and potential restrictions on the companies with Trump administration officials. People familiar with the matter said, OpenAI offers some open-weight models, and Altman, its CEO, said last year that he regrets the company's past stance against the tools. The Trump administration is considering adding Chinese companies to a trade blacklist that would effectively prevent many U.S. businesses

from using the models. Company critics, including some administration officials, see Anthropics warnings as part of an effort to hinder its competitors, given that many such open models could undermine its attempts to maximize revenue as it plans a public listing as soon as the fall. Kevin Bryan, an associate professor at the University of Toronto, specializing in innovation, defended Anthropic and said the fears about the risks of frontier intelligence being open source are real and widely shared. Anything that is open source is immediately jailbroken, he said. Once the models out, that's it. Some critics argue Anthropic itself has built models through a process that resembles distillation. It's interesting that they're against distillation when they distill the entire internet without

paying royalties, said Mark Suman, co-founder and CEO of Maple and AI productivity tool startup. I just think it's odd that they are now turning around and telling others not to do what they did. Hours after the letter from companies supporting open-weight models became public OpenAI signed. Alphabet's Google endorsed the letter shortly after leaving Anthropic as one of the only notable AI companies not to back it. More than 70 Silicon Valley firms have now added their names. Technologists and policymakers say they are also concerned about the AI race turning into a duopoly with Anthropic and Open AI essentially emerging as dominant players that control the market. Some founders and startup executives said Anthropics release of Claude Design was a wake-up call across Silicon Valley.

The product release led some founders to conclude that Anthropic might launch tools to compete with them and take their customers. They turned to cheaper open weight models in earnest after that, particularly ones that weren't built by U.S. Frontier Labs, end quote. This episode is brought to you by Accenture. When your advertising operations fall out of sync, everything else follows. Spotify and Accenture are working together to reinvent the rhythm of ad sales, Using automation, analytics, and smarter workflows to simplify campaign delivery and access better data across the business. The result? Less time spent on operations, more time connecting brands with the moments and fandoms that matter most.

Learn more at Accenture.com slash Spotify. When critical company knowledge isn't documented, there's a major ripple effect. Work becomes inconsistent. Tools don't get adopted and knowledge walks out the door when someone leaves. Thankfully, our sponsor, Scribe, was a major ripple effect. built to fix that. Their workflow AI platform is trusted by nearly half of the Fortune 500 to capture workflows in real time. Here's how it works. You turn on the Scribe browser extension or desktop app, do a process as you normally would, and Scribe will build a guide as you go.

It automatically redacts sensitive information and even suggest improvements to your existing workflows. To see what Scribe could look like for your org, head to Scribe.com. Rideholm and mention Rite Home for your first month of Scribe capture free on select plans. That's S-C-R-I-B-E dot how slash ride home. If you only know a little bit about Bitcoin, that's okay. If you've been curious about Bitcoin but haven't made the jump yet, Cash App makes it easy. You can set up automatic purchases with zero fees or buy larger amounts, also with zero fees.

Start small or go bigger. It's designed to be simple either way. For a limited time, new customers can get $10 added to their balance. Just use code Bitcoin 10 when you sign up. And don't forget this part, send at least $5 to a friend in the last. the first two weeks. Terms apply. Cash app is a financial services platform, not a bank, banking services provided by Cash Apps bank partners, Bitcoin services provided by Block Inc. brand. For additional information, see the Bitcoin disclosures at cash.com.

slash legal slash podcast. Now, the final bit of context from a substack called the Algorithmic Edge, that is the best summation I've seen about possible motivations among the biggest players right now. quote. What if I told you that Google is out of the AI race? It didn't happen because Google lost, though, despite an underwhelming I-O event earlier this year, several heavyweight departures and a slower model release pace than its competitors. Google is still alive. It is out of the race because it withdrew. Weird, right? Why would Google do that willingly? To answer that, I'm going to tell you the story of why and how Google left the AI race that OpenAI and Anthropic are betting everything on. Here's my hypothesis stated plainly. Google DeepMind

CEO Demis Hasabas, a pioneer of AI, as we understand it today, doesn't think that automating AI research with coding agents, AI systems that can program better AI systems themselves, is the correct approach to artificial general intelligence. Google's deep-mind leadership sees Open AI and Anthropics bet on recursive agentic programming at best as an off-ramp and at worst as a dead end. Hasabas is betting on something else. World models, models that can understand and simulate the real world, not just predict the next token. And he's steering Google accordingly. The AI race is actually a race between two theories of intelligence and also between two kinds of companies. Startups that need AI to become a

business and an incumbent whose existing business can subsidize AI for years. As we will see, Hussabas's decision puts Google in a life or death situation. He could make Google the absolute leader or an irrelevant laggard. That said, to write Google off the story would be a serious mistake. I'm going to explain why I think Google has a unique chance here. The first piece of the puzzle is that Anthropic CEO Dario Amadai and OpenAI CEO Sam Altman think that AI models will soon be autonomously improving themselves. This is known as recursive self-improvement or RSI. Model A makes a better model B, that makes a better model C, etc., and that this is the key

to achieving AGI. The best approach to pursue this goal is also the simplest. Build data centers and use brute force on algorithms that can be scaled that way, like learning and search. That's the fundamental insight of Richard Sutton's bitter lesson. which both OpenAI and Anthropic hold sacred. If new architectures or algorithms are needed, those better models equipped with a lot of compute will find them somehow. That's the theory, at least. What does this creed look like inside the lab? We only need to read their testimonies.

You have engineering staff at both Anthropic and Open AI admitting that they barely write any code anymore. It is swarms of coding agents, Claude Code and Codex that do that. Employees have become managers of agents rather than developers and researchers themselves. Claude Code writes Claude Code and wrote Claude Co-work, GPD 5.6 sole autonomously post-trained GPT 5.6 Luna. They say straight away that RSI is coming. Human supervision is still required for many things. We're in the final moments of the human in the loop stage. But one prediction is that by 28, there's a 60% chance that RSI is achieved. That's what Anthropic co-founder Jack Clark wrote in a newsletter in May. Others predict RSI even sooner.

Being six years younger than opening eye, Anthropic had one shot at success, and they hit the bull's-eye. An unbreakable focus on code plus agents plus enterprise secured a steady climb to the top. It's been a crazy streak of success, from revenue to valuation, to popularity, to getting ahead of larger and older companies, passing through several back and forths with the U.S. government as well. Open AI is a different story. It skyrocketed to fame with Chad CTP, and then it had to deal with the success disaster that followed. Too many expensive free users not willing to pay. At the time of writing, the revealed preferences of U.S. households indicate that only 2.2% are willing to pay for an AI subscription. Open AI realized that Anthropic was starting to show the signs of

acceleration consistent with Proto RSI. They care about money, sure, but Altman wouldn't make a company-wide restructuring like he recently did for a short-term trade. Thus, their recent full operational overhaul intended to be a long-term investment. Open AI can't afford to let Anthropic win in the enterprise market, but it can't afford someone else achieving RSI-enabled escape velocity. OpenAI pivoted fast. It doubled down on Codex, Agents, Enterprise, and ultimately on the road to recursive self-improvement. OpenAI is doing so well now that Insiders confidently claim that Codex is actually better than Claude Code. Codex actually grew well beyond expectations after the release of chat GPT work, and it's up to 10 million users, as Bloomberg last reported. What I know is that those left out of the first stage of the exponential are going to be out of the game, and as far as we can tell from public information, that would be Google. Anthropic made this strong claim far back in April 2023. We believe that companies that train the best 2025-26 models will be too far ahead for anyone to catch up in subsequent cycles. This is what I mean when I say escape velocity. It's the speed you need to escape the gravity field of the planet without falling back to the ground. Once you get there, your next model release is accelerated thanks to the previous models involvement in training and post-training. At the time of writing, everything Anthropic and opening I argued in 2023, 2024 has become true. All of the above happened gradually and then suddenly in the first few months of 2026.

Where's Google in this entire story? Can you see them? Yeah, they have Gemini 3. Something and anti-gravity and something called Omni. More on that later. But they're essentially nowhere to be found at the frontier of the code plus agents plus Enterprise plus RSI race. Why? End quote. He goes on to discuss Google's bet on world models and why basically only they can make that bet at this point. It's an 8,000 word piece, so click through if you want to hear that part of it. But if you put all these segments together from today's show, I think you've got a better sense of why everybody might be behaving like they are.

People are in various levels of freak out, but for different reasons. Talk to you tomorrow.