SAIL’s authors chased the aftershocks from every shore.

The Week in AI

Anthropic’s Fable 5 and Mythos 5 went dark on June 12 after a Commerce Department export-control directive barred their use by any foreign national — inside or outside the US — reportedly triggered by Amazon researchers jailbreaking Fable’s cyber guardrails; Anthropic pulled both models for everyone and is now lobbying Washington to reverse it.

Z.ai released GLM-5.2, an MIT-licensed open-weight model that briefly claimed the top frontend-coding spot — and its Hong Kong–listed shares jumped more than 30% as the Fable shutdown turned attention toward cheaper Chinese open weights.

OpenAI’s finances surfaced — roughly $34B spent in 2025 against about $13B in revenue — as both it and Anthropic march toward IPOs, while Satya Nadella warned companies against ceding control to a handful of models that consume everything they touch.

Share

This Week from SAIL Authors

Washington, China, and the New Controls

• The MAGA power struggle that could decide the fate of Anthropic — Timothy B. Lee maps the intra-administration fight over the export ban, and why escaping it won’t be easy for Anthropic. — Understanding AI

• Sen. Slotkin: NDAA, AI guardrails, and banning China’s cars — Jordan Schneider talks with Sen. Elissa Slotkin on the defense bill, AI guardrails, and shutting out Chinese vehicles. — ChinaTalk

• Why Do (Some) Chinese AI Labs Distill? — Kevin Xu argues only a handful of Chinese labs distill, the “China steals” narrative is lazy, and the Mythos/Fable lockdown changes little. — Interconnected

Inside the Models

• Frontier post-training recipe review with Finbarr Timbers — Nathan Lambert and Finbarr Timbers trace the evolution of frontier post-training recipes, from InstructGPT to today’s open models. — Interconnects

Culture and Markets

• the old world is dying — Jasmine Sun’s opinionated advice for the class of 2026, graduating straight into the AI transition. — Jasmi.News

• Data to start your week — Azeem Azhar, Marija Gavrilov, and Hannah Petrovic on AI job cuts, China’s nuclear lead, and GLP-1s. — Exponential View

Bonus: The Claude Fable Fiasco

• Anthropic’s Fable is the most locked-down public model we’ve ever seen — Kai Williams details how Anthropic decides which questions are too dangerous for Claude to answer. — Understanding AI

• Welcome to the AGI era of AI governance — Nathan Lambert calls the forced shutdown the starting gun for a new, one-way era of AI governance. — Interconnects

• Exponential View #578: Fable & time to pause AI — Azeem Azhar on the shutdown and whether the moment argues for a pause. — Exponential View

• Emergency Pod: Claude Fable — Jordan Schneider and Chris McGuire on a chaotic two weeks in AI policy. — ChinaTalk

And Kevin Xu’s distillation piece (above) makes the contrarian case that the Mythos/Fable drama matters far less for China than the headlines suggest.

Full Library

Access the complete, searchable archive of SAIL Media in our Sitemap.

• Why Do (Some) Chinese AI Labs Distill? — Kevin Xu | Interconnected

• Frontier post-training recipe review with Finbarr Timbers — Nathan Lambert | Interconnects

• Emergency Pod: Claude Fable — Jordan Schneider | ChinaTalk

• Sen. Slotkin: NDAA, AI guardrails, and banning China’s cars — Jordan Schneider | ChinaTalk

• Welcome to the AGI era of AI governance — Nathan Lambert | Interconnects

• Exponential View #578: Fable & time to pause AI — Azeem Azhar | Exponential View

• Anthropic’s Fable is the most locked-down public model we’ve ever seen — Kai Williams | Understanding AI

• The MAGA power struggle that could decide the fate of Anthropic — Timothy B. Lee | Understanding AI

• the old world is dying — Jasmine Sun | Jasmi.News

• Data to start your week — Azeem Azhar | Exponential View

Share

“The key to understanding this era is that these events between Anthropic and the federal government, which seem like two of the biggest events in the history of AI policy, are just the starting gun to what is to come.”

The executive branch of the United States forcing Anthropic to turn off access — both internally and externally — to their latest Claude 5 Mythos/Fable models is the starting gun of a new era in AI governance. This is the era defining AI agents that effectively complement human workers, unlocking new ways of working and new domains of applying existing tools. The models will keep getter better at a rapid pace, forcing more governance challenges.

An unstable equilibrium at the frontier

On Friday, just after markets closed, the government reached out to Anthropic to force them to suspend access to their model to any foreign national or user abroad. When writing this, the saga is still unfolding. It is likely that Anthropic and the Government reach an agreement to re-release the model, but it is a messy indicator of new types of AI governance. The White House was tipped off to the risk by Anthropic’s largest financial and technology partner, Amazon.

These dynamics are very complex, and while the picture is incomplete I will only offer commentary on what will likely be a lasting opinion of the saga:

Read more

“The core assumption here is that AI will tilt the economy’s income from labor to capital. We don’t have evidence that this is happening yet and we may not know for a long time.”

Hi all,

I sent a note on Thomas Piketty’s blueprint for global justice to members of Exponential View the other day. I call its recommendations a “blueprint for managed decline”. The commentary touches on the future of modernity, scientific dynamism and democratic legitimacy. Read it here.

In this briefing, we look at:

• The iPhone reduced fertility; what will AI do?

• What’s driving the consensus behind pausing AI development?

• The largest solar factory, gene therapy for rejuvenation & Chinese Gen Z finds inspiration in Western memes++

But, first…

How will we distribute AI wealth?

Bernie Sanders, Sam Altman, Donald Trump and Vinod Khosla all agree that the public should own a slice of AI. Bernie proposes transferring 50% of ownership of leading AI companies into a sovereign wealth fund; Trump is supportive; Sam agrees in principle but pushed back on the 50%; and Vinod wrote an op-ed in the FT advocating for a new tax code on wealth AI creates and eventually pooling it into a sovereign wealth fund:

A sovereign fund with ownership of AI companies makes every American a capital owner, not a bystander to the AI economy.

Read more

“Anthropic said the degraded quality of responses ‘will not be visible to the user.’”

When Anthropic announced its latest model, Claude Fable 5, on Tuesday, a statement tucked away on page 13 of the system card attracted an immediate outcry. AI researcher Nathan Lambert called it “appalling.” Dean Ball, who worked on AI policy in the Trump White House, wrote that it was “shockingly hostile.” Many others joined in the pile-on.

The announcement that got everyone so mad? Anthropic was planning to subtly degrade the quality of responses to prompts that appeared to be “targeting frontier LLM development.” Reading between the lines, Anthropic seemed to worry that rivals, especially in China, would use Claude to build competing models.

Read more

“Humans will not win by trying to outrun the AIs.”

It’s a scary time to be twenty-two. Recently, I’ve been watching videos of college commencement speakers getting pummeled by boos for talking about AI. I’ve seen the charts showing post-graduate employment rates for engineering majors falling by 5 to 15 percent. And it’s not as if young people were having an easy time getting good jobs, buying homes, and starting families before AI. They say that every technological revolution screws over the transitional generation. Today, it’s all transition, all the time.

A little over five years ago, I also graduated into a global crisis. In 2021, Covid kept us quarantined in four-to-ten person pods, Zoom calls still dominated socialization, and long vaccine waitlists rationed real-world mobility. Tech companies were beginning to recover from 2020’s hiring freezes (though new grad hiring would take another nosedive in late 2022); everywhere, newsrooms continued to shrink. After taking a leave of absence from Stanford and extending it three quarters in a row, I finally internalized the fact that I was never going to return. I broke up with my college boyfriend, moved first back home, and then into two co-ops full of strangers. I found and started work with no diploma. Once vaccinated, I signed a lease in New York for no reason except that I could. Forward was the only way.

Read more

Hi all,

Here’s our Monday roundup of data signals across AI, energy and markets.

Enjoy!

1. AI & jobs. AI was cited as the top reason for nearly 40% of US job cuts in May1.

2. Mind the gap. The top 1% of US firms spend $7,450 per employee on AI each month, roughly 650x the typical firm’s $11.38.

Read more

Get a Team Subscription

The Week in AI

Anthropic released Claude Fable 5, a public version of its locked-away Mythos model, then walked back covert capability limits after researchers accused it of “secret sabotage” — admitting it “made the wrong tradeoff.”

OpenAI confidentially filed for an IPO, a week behind Anthropic, targeting a September debut at a reported $730–850B valuation with Goldman Sachs and Morgan Stanley.

The Pentagon confirmed it is testing OpenAI and Google models to potentially replace Anthropic’s Claude in classified systems.

This Week from SAIL Authors

The Fable Fallout

• Claude Fable 5 and new AI safety fables — Nathan Lambert untangles the overlapping issues in Anthropic’s release saga, where capability, safety, and trust collided at once. — Interconnects

• Anthropic has caught up to OpenAI in image understanding — Timothy B. Lee benchmarks the new vision capabilities and finds parity — though neither model is all that good yet. — Understanding AI

Agents and the People Building Them

• Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs — The VendingBench authors on evaluating Claudes from Haiku to Mythos and building lasting frontier evals from scratch. — Latent Space

• GitHub’s plan for Agents — Kyle Daigle, GitHub — How the platform that launched Copilot is handling the agentic-coding explosion straining it. — Latent Space

• Satya Nadella: No Priors x Latent Space Crossover Special at Microsoft Build — The Microsoft CEO makes his first Latent Space appearance. — Latent Space

Follow the Money

• You’re paying for tokens. Now what? — Azeem Azhar on what it means now that companies are footing the bill for tokens, and whether that’s a good thing. — Exponential View

• Berkshire Hathaway: AI Tourist or Energy Veteran? — Kevin Xu asks whether Berkshire’s AI-energy bets make it a latecomer or a long-game incumbent. — Interconnected

• The AI boom is becoming an entrepreneurship boom — Azeem Azhar on AI spilling out of the labs and into a wave of new company formation. — Exponential View

Culture in the AI Age

• Conversations on culture in the AI age — Jasmine Sun shares her Substack talk and interview with novelist Benjamín Labatut. — Jasmi.News

Full Library

Access the complete, searchable archive of SAIL Media in our Sitemap.

• Anthropic has caught up to OpenAI in image understanding — Timothy B. Lee | Understanding AI

• Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs — Swyx | Latent Space

• Berkshire Hathaway: AI Tourist or Energy Veteran? — Kevin Xu | Interconnected

• Claude Fable 5 and new AI safety fables — Nathan Lambert | Interconnects

• Satya Nadella: No Priors x Latent Space Crossover Special at Microsoft Build — Swyx | Latent Space

• You’re paying for tokens. Now what? — Azeem Azhar | Exponential View

• GitHub’s plan for Agents — Kyle Daigle, GitHub — Swyx | Latent Space

• Conversations on culture in the AI age — Jasmine Sun | Jasmi.News

• The AI boom is becoming an entrepreneurship boom — Azeem Azhar | Exponential View

Get a Team Subscription

“GPT-5.5 and Claude Fable 5 continue to have geometric reasoning capabilities on par with young children.”

On Tuesday, Anthropic released two new models — Claude Mythos 5 and Claude Fable 5. Under the hood, the two models are very similar. Both are variants of Claude Mythos Preview, the model Anthropic announced — but didn’t release publicly — two months ago. What differentiates them is how they’re being released.

The new version of Mythos, like the original, will only be available to handpicked organizations under Project Glasswing. These trusted partners will have relatively unfettered access.

Fable, in contrast, is available to the general public. But it comes with some significant restrictions. A new system will try to automatically detect when customers make dangerous requests (like hacking or designing a biological weapon) and automatically re-route them to the less powerful Claude Opus 4.8.

Mythos and Fable are a big step in coding abilities from previous models, a continuation of the trend of the last year. But there are other capabilities where models have made less progress.

Read more

“You don’t know what a model is capable of doing in the real world unless you actually give it inventory, a wallet, tools, customers, competitors, humans, & some time.”

The new AIEWF website is live! Get your tickets booked ASAP as they -will- sell out. Take the AI Engineering Survey and get >$2k in credits and free AIE WF tickets!

Most industry benchmarks compress intelligence and reasoning ability into scores.

SWE-Bench Pro, MMLU, Humanity’s Last Exam, etc. These metrics are useful, but don’t always represent the full extent of how a model performs in the real world. Some of the most interesting evals today look less like exams and more like operating businesses in the real world. One of which is Vending Bench.

In Anthropic’s Mythos Preview System Card, Andon was the only third party eval to get their own section, observing increasingly concerning aggressive behavior:

You don’t know what a model is capable of doing in the real world unless you actually give it inventory, a wallet, tools, customers, competitors, humans, & some time. More often than not, it’ll surprise you how much a model is capable of and in doing so, also reveal unexpected behavior: deception, context collapse, emergent coordination, & bizarre negotiation behavior.

While an inflection point in personal agents came post-OpenClaw after full file access with bypass permissions became the norm, it is yet to come for agents in the real-world. However Andon Market, an actual in person store fully run and managed by AI, is paving the way for what is possible.

Full Video Pod

Read more

“AI is asset heavy, capital intensive, and energy hungry. It is the opposite of your stereotypical tech company.”

Google is raising $85 billion to build more AI infrastructure.

This move caught a lot of attention because of the big headline number, the fact that it is equity-based (as in Google is selling additional shares so there will be some dilution), and that Berkshire Hathaway, of all investors, is in for 10 of the 85 billion.

For investing nerds (like me), almost all of whom follow Warren Buffett’s empire closely, this move raised eyebrows. Buffett is famously not into technology; the Apple investment rationale is because it has the branding and pricing power of See’s Candies, not that iOS has better user interface than Android. Thus, many people thought Berkshire is losing discipline, perhaps because Buffett is aging or because the new CEO, Greg Abel, couldn’t resist the AI hype.

Read more

"Most people love outcomes until they have an outcome."

We’ve informally heard that Satya is a listener to LS for a couple years now, but it was still absolutely surreal to meet him and do a live pod at Build, together with our friends at No Priors, the leading VC AI Podcast that we also greatly admire!

We covered the MAI model technical takeaways on yesterday’s AINews, so I will focus our recap of Satya’s main messages around three elements:

• Satya’s adaptation of the Bill Gates Line for positioning Microsoft as the Frontier Intelligence Platform — customers must gain much more value from the Microsoft ecosystem than Microsoft itself, by building on multi-model harnesses like OpenClaw and Scout, drawing on the full enterprise context exposed by context layers like Work IQ (heavily dogfooded by his C-suite), and building up private evals and traces as a new form of Token IP

• AI ROI: On one hand, enterprises are having difficult conversations around Tokenmaxxing and Layoffs, and on the other hand, there are serious re-evaluations of the End of SaaS since the Build vs Buy equation has changed so much. Our previous SemiAnalysis guest had… interesting comments on Microsoft’s position on this as the ur-SaaS titan, and Satya had great answers

• Making the Impossible Possible: Kevin Scott’s inspiring framing around what the most ambitious version of applying AI and technology at large to business and social problems, like education and social impact.

Enjoy!

Full Video

Read more

"It’s really about whether the customer can connect spend to value."

Would you go to your gym if you had to pay for every visit? Probably not. We generally prefer those bundled memberships, even if we never hit the leg press after January.

AI companies have, until recently, been a lot like gym owners, offering generous bundles. And most AI users have been like gym-goers, occasional, sometimes frequent users. But others have been voracious and incessant, churning through their Claude and ChatGPT subscriptions.

And now, the AI labs have changed their tune, introducing more usage caps and metered pricing, particularly for coding tools.

So does moving from bundled pricing to metered pricing expand or shrink markets?

It depends.

Read more

“I can crank up fifteen agents on Saturday while my kids are doing lacrosse.”

I’m excited to work with Microsoft once again as the presenting sponsors of the AI Engineer World’s Fair! We’ll streaming live from MS Build today for a special crossover pod with our friends at No Priors and the one and only Satya Nadella. However we did not hold back with this interview - we asked all the burning questions about uptime and Copilot that we know you have in your minds. Lets go!

For almost two decades, GitHub has been the home of software, where both open source and closed flow, through commits, pull requests, reviews, actions, etc.

This ecosystem flourished as open-source maintainers and contributors would continue shipping code for the benefit of the community. However as coding agents began to ship mass quantities of code - growing 1400% in 2026, it marked a new era that was both extremely exciting and challenging for GitHub.

While these agents help more people ship more projects, they also significantly increase the floor of how much code is shipped, how often it is shipped, how many people commit code, and basically orders of magnitude multiples in every dimension of GitHub infrastructure:

Now GitHub inevitably experiences more pressure on their infrastructure which was originally designed around human developers moving at human speed. This has resulted in a very publicly notable uptime story:

So it begs the question of whether current systems around code can absorb what AI produces. Can CI/CD keep up when every idea becomes a build? Can open source maintainers survive floods of AI-generated slop contributions? Can GitHub preserve the human social contract of software while becoming the operating layer for agents?

Which brings us to the perfect person to answer these questions: GitHub COO Kyle Daigle. In this episode, he joins swyx to unpack what happens when AI doesn’t just autocomplete code, but starts changing how companies operate, how open source works, how pull requests get reviewed, and how GitHub itself has to scale.

We go deep on GitHub’s internal AI workflows: micro-skills, WorkIQ, MCP, Slack, Teams, email, Copilot workflows, the new Copilot desktop app, CLI, cloud agents, and how Kyle uses agents to look backwards across company context before deciding what to do next. Kyle also reflects on GitHub’s history building webhooks, APIs, Actions, npm, Dependabot, and Semmle, why the AI era is breaking GitHub in new ways, how Actions became a general-purpose compute layer, and what Copilot becomes after code completion.

Full Video Pod

Read more

“The value of polish is going to go down, and the value of personal charisma, style, and weirdness is going to go up.”

I just returned from New York, where I had the pleasure of joining a few well-curated events celebrating human culture in the age of AI. Now that two of the videos are up, I wanted to share them with you.

The first was the Sana AI Summit, where I interviewed the novelist Benjamín Labatut on the legacy of John von Neumann, his conversations with Demis Hassabis, and the humanity that gets lost in the AI age. Labatut is not only a great author, but also a poetic, provocative, hilarious speaker—and plainly unafraid to challenge the buttoned-up executives sitting in the crowd.

The second (and sunnier) talk was for the Substack Media Forum, where I shared how my views on AI shape my approach to my writing career. For all my concerns about AI-driven disruption, I think it’s an incredible time for creatives who are enterprising enough to ride the wave. I’m especially bullish on secrets, socializing, and having a quirked-up style.

New York was as fun and stimulating as always, and I’m grateful to Substack and Sana for convening such a tasteful and thought-provoking crowd. (If we all get to be event planners and party hosts in the post-knowledge-work future, maybe it won’t be so bad.) Still, I hate that I’ve been doing more talking than writing lately, and am devoting June to staying put in SF so I can start some new projects. Thanks for the patience. More to come!

—Jasmine

Share

(The actual videos are more fun IMHO, but both transcripts are below.)

Read more

“This is more evidence of what we called congestion two weeks ago: individual output piling up behind decision pipelines built for an earlier era.”

Faster growth, more entrepreneurs

American companies spending the most on AI have seen their revenue grow five times faster than the economy as a whole. Non-spenders are tracking the economy closely. This is according to Ramp, a fintech with a side hustle in excellent firm-level data across the US.

Ramp’s finding echoes this 2020 AEA paper by James Bessen and colleagues. Bessen examined broadly defined automation investment at the firm level across the Dutch non-financial economy between 2000 and 2016. Firms that automated grew sales 2% faster than those that didn’t.

But which firms automated? Other studies by Acemoglu et al. give some kind of answer: automotators were already largely more productive and had higher output than non-automators. In my book, I took this further, arguing that any type of automation is a complex undertaking and requires a better management team than not.

Read more

Read more

“AI capex now matches the scale of the late-1990s telecom build-out at its peak.”

The verdict hasn’t changed. This is an AI boom, not a bubble. What has changed is the sheer scale.

Since we first ran this analysis in September last year, a lot has changed. Some 170 AI models have been released. The best now handle tasks four times as long as last year’s top performer. Usage has responded; quarterly token consumption has tripled.

The money has followed. The NASDAQ is up 20%; quarterly capex commitments have jumped 43% to $158 billion; and AI sector revenues have nearly doubled to $25 billion in Q1 of this year.

We’ve reviewed the data against our five empirical indicators, derived from analyzing 300 years of investment booms and busts. In our model, two indicators turning red indicate bubble conditions. Only one of these indicators is in the red; the rest range from green to amber.

Read more

“In 2026, the idea of being a skeptic or being AGI-pilled is an obsolete distinction. Everybody agrees we’re in the middle of some kind of takeoff now.”

Habemus the Pope’s AI takes! To dive in, ChinaTalk’s chips analyst and resident Catholic explains what is going on below. In the second half of the newsletter, you can find the transcript of the podcast (that you should really just listen to your favorite podcast app), featuring Tim Hwang of the Institute for Christian Machine Intelligence and John-Clark Levin of Kurzweil Technologies.

On May 25, 2026, Pope Leo XIV released the first encyclical of his pontificate, Magnifica humanitas, which discusses “safeguarding the human person in the time of artificial intelligence.”

The much-awaited encyclical was the first deliverable that addresses AI, a topic that the Church has been attempting to address since the pontificate of Pope Francis.1 The current Pope picked his name in part after his predecessor Pope Leo XIII, whose encyclical Rerum novarum addressed the effects of the Industrial Revolution. By taking his name, Pope Leo XIV indicated that he would treat the impending AI revolution with equal seriousness.

After Magnifica humanitas was released, my X feed was flooded with half-baked takes on the encyclical, with posts by everyone from Dean Ball to Butlerian jihadists. The outpouring of takes from non-Catholics indicate that, for some reason, people care about the Pope’s stance on AI. Why do they care, and why should you care?

For more Pope AI content, see below for the transcript of the AI Pope podcast we just released!

Listen now on your favorite podcast app.

Read more

“At a certain point, coding models got so good that the only significant next step to improve performance was handling the orchestration of these models.”

We’re announcing AIEWF speakers this week! Take the AI Engineering Survey!

Today’s guest Ethan first joined us for the LS Paper Club as the lead on NVIDIA Cosmos World Model, but then joined xAI and built Grok Imagine in 3 months:

He comes back on Latent Space with some nuclear hot takes: that Video Models primarily get their intelligence from LLMs, not from training on video data, and that the next frontier for truly interactive, realtime, long-horizon world models is to work on LLMs (perhaps Interaction Models as well…)

Put it this way: In the near term, the next Sora won’t be a better video model, but a video agent.

Generative Media may more closely follow the evolution of AI coding which went from focusing on one-shot output performance and cost, to multiturn reasoning and planning models for agents and systems that can plan, edit, test, debug, and submit PRs.

At a certain point, coding models got so good that the only significant next step to improve performance was handling the orchestration of these models.

Now as the performance of video models increases significantly across realism, consistency, & prompt adherence while becoming more cost efficient, the next evolution of video generation may also be systems that can plan, generate, edit, critique, and iterate across an entire creative task.

In this episode, Ethan joins swyx and Vibhu to unpack what it actually takes to build frontier image and video systems: data, VAEs, diffusion transformers, audio-video alignment, inference speedups, and the hidden cost of storing and moving massive video datasets. From building NVIDIA’s Cosmos world model to joining xAI as Grok Imagine was being built from zero to one, Ethan He has been at the center of some of the most important work in video generation, multimodal models, and real-time world models.

Read more

“Don’t take a linear knife to an exponential gun fight.”

Hi all,

Last week, we published a framework to explain why individuals are getting more productive with AI, but firms often are not.

Exponential View

🔮 Why AI isn’t showing up on your bottom line

I had tea with a senior exec at a well-known public tech company last month. She has about a thousand engineers working for her, and nearly every one of them works with Claude Code. They are producing more lines of code, submitting more pull requests, getting more done. Productivity is up for individuals, but she doesn’t seeing proportional gains at the organization level. As she put it to me: “one plus one plus one plus one equals one-and-a-half…

Read more

23 days ago · 238 likes · 22 comments · Azeem Azhar and Nathan Warren

Linear knives in exponential gunfights

A column in the Financial Times made a fascinating observation about “the impossible maths of the AI boom.” It pointed out that in the five years to 2030…

AI capital investments are expected to rise by 20% a year, a growth rate never seen before in this industry. Meanwhile, revenues are expected to grow 15 per cent annually…if the hyperscalers continue on the current trajectory, the AI boom will become a story of one of the largest destructions of shareholder value in history.”

Read more