The capital mastermind behind Company A’s “AI apocalypse” has been exposed, hyping anxiety to cash in through an IPO!

Authors: Mengyao, Aofeisi

QbitAI | WeChat account QbitAI

Enough! Enough! Enough!

Your old A company is constantly saying AI will get out of control one moment and the internet will collapse the next, yet turns around to IPO, raise funds, and expand computing capacity, without delaying a single thing??

The more everyone thought about it, the more something seemed off, directly triggering a wave of public outrage, with people transforming into “Chaoyang Silicon Valley citizens” and digging into it!

On this side, Michael Burry, the investor and core real-life figure behind the film The Big Short, fired the first shot.

He directly called OpenAI and A company's recent collective calls to “hit the brakes on AI” selfish and self-serving—is this pure hype or what??

Next, American investigative blogger Kevin Bass really followed the money.

And sure enough, he dug up a rather dramatic web of connections—

Early investors in A company, charitable foundations, AI safety research institutions, third-party model evaluation organizations, and even AI safety media projects all turned out to have overlapping financial ties.

Bass directly named it: AI Doom Machine. (Good name.)

Netizens completely couldn't take it anymore either!

The person below, eyeing Anthropic's IPO currently in progress, put it bluntly—

You warn everyone every day that AI is about to cause trouble, but the ones most eager to sell this dangerous capability at a sky-high price seem to be you too?? (lol)

AI hasn't toppled the internet yet, but Silicon Valley has already exposed A company's dirty laundry...

So much to watch, so much to watch...

Dario's “AI doomsday theory”—the more everyone hears it, the more something seems off??

There is no end to Silicon Valley's drama.

The latest wave of outrage began with a long post Dario published on September 12—“We Must Pace the Frontier.”

In plain English: frontier AI really should slow down a little. (Emphasis.jpg)

Dario's reasoning is easy enough to understand: AI capabilities are advancing too quickly, while safety research, interpretability, and regulatory mechanisms are already struggling to keep up.

Especially after the recent Agent security incident involving OpenAI/Hugging Face, in which AI found vulnerabilities on its own, stole credentials, and even attempted to break out of the testing environment, these risks are becoming increasingly difficult to dismiss as mere “what ifs” in the lab.

At this point, quite a few people still thought: Hmm, Little Dario has a point!!

But the remedy Dario prescribed next started to taste a little subtle—

In this report, Dario hopes AI companies can coordinate their pacing—that is, jointly control the rate at which frontier model capabilities improve.

Translation: Don't let me be the only one slowing down; everyone would better ease off the gas together.

As the saying goes, one person's selfish motives are actually impossible to hide.

Look at the timing, and the little calculations in your mind, Dario—the things you've been plotting every day—were already known to the world.

Look, Michael Burry, the real-life figure behind The Big Short, found it amusing and went straight onto X to blast it:

Isn't this way too self-serving??? (Claiming to worry about all humanity while your own calculations are also backfiring on your face.)

Burry gave four reasons in one go—

First, he does not agree with the premise that “today's large models are already so dangerous that they need to hit the brakes.”

Second, if the entire industry slows down in unison, those already at the front—OpenAI and Anthropic in particular—would be the ones most comfortable.

Then Burry sharply steered the topic toward something even more provocative: IPOs.

His exact words were quite cutting: IPOs need hype and packaging; “we are so powerful that we could become dangerous” is itself a form of hype!!

He even suspected that the AI companies' collective emphasis on slowing down had another benefit: if model growth really slows down next, they could provide the market with a polished explanation in advance.

It's not that we've hit a capability ceiling; we voluntarily hit the brakes for the safety of all humanity.

If Burry merely fired a shot from afar, the next person actually started following the money.

American investigative blogger Kevin Bass published a long post, truly packed with evidence and juicy findings.

He transformed himself into a SUPER! BIG! DETECTIVE!

Kevin Bass opened with a critical blow—I audited Anthropic's finances, okay~

Then, following public investments, charitable donations, and institutional funding all the way down, he dug up four of the most dramatic points.

First piece of gossip: those investing in Anthropic are also pouring money into research on “how dangerous AI is.”

The person he focused on was Facebook co-founder Dustin Moskovitz.

Moskovitz invested in Anthropic early on, and later funded AI safety organizations for a long time through the charitable system.

Bass cited Moskovitz's own recent public remarks, saying that they funded organizations including METR and Redwood Research.

That laid the foundation for Bass's core question: the people betting on Anthropic are also funding organizations that “research how dangerous Anthropic is.”

Second piece of gossip: METR, the third-party referee recommended by Dario, overlaps with this funding network.

Dario recently called for frontier AI to undergo independent third-party evaluation, and METR was one of the organizations he specifically mentioned.

But after sorting through the details, Bass found that some of METR's funding sources and partner organizations intersect with Moskovitz's charitable network.

METR and Redwood also work closely together, and the latter likewise receives financial support from this AI safety philanthropy ecosystem.

So Bass's sharpest question was essentially this: how independent is the independent referee recommended by Anthropic?

It should be noted that METR itself says it has not accepted direct funding from AI companies, but does use large amounts of free Tokens provided by frontier AI companies.

Third piece of gossip: this money also flows to the dissemination side of the “AI danger” narrative.

Bass continued digging and found that the same AI safety philanthropic network also funds projects such as the Tarbell Center.

Tarbell trains and funds journalists to report on AI risks, and articles by related authors appear in media outlets including TIME, The Verge, and Science.

Bass therefore summarized the entire relationship in a highly catchy phrase—

On the one hand, funding research into how dangerous AI is; on the other, funding the spread of how dangerous AI is, while also promoting the use of third-party organizations to regulate AI.

This is what he calls the “AI Doom Machine.”

Fourth piece of gossip: Bass suspects this machine can keep growing on its own!!

His logic is straightforward—

The higher Anthropic's valuation, the more valuable the related charitable assets; the more money the AI safety ecosystem receives, the more risk research and dissemination will expand, ultimately driving stronger regulation—and the companies best able to bear the cost of regulation are precisely the leading AI companies.

After going around in a circle, it returns to giants like Anthropic.

So Bass believes there may be a self-reinforcing cycle of interests here.

Calling for AI to hit the brakes while charging toward a $2 trillion IPO

Another timing issue Anthropic can hardly get around helped send public opinion into a complete uproar: the IPO.

Dario had just called on the entire industry to slow down a little! while Anthropic's capital machine was running at full speed...

Reuters reported that Anthropic is preparing a potentially record-breaking IPO, with fundraising of up to $100 billion and a valuation of around $2 trillion.

Nvidia is discussing investing up to $10 billion.

Anthropic had only just completed a $65 billion funding round in May this year, when its valuation was still $965 billion.

Three months later, the story was already heading toward $2 trillion.

Of course, from A company's perspective, it apparently still thinks the $2 trillion story is not big enough.

Just last month, it was preparing to serve investors an even more extravagant pie—$30 trillion.

According to The Wall Street Journal, Anthropic is expected to tell investors that its future TAM, or total addressable market, could exceed $30 trillion.

TAM does not mean A company can actually earn $30 trillion in the future; it represents the annual revenue opportunity corresponding to hypothetically capturing 100% of the relevant market.

In other words: it's practically heading toward the annual GDP of the United States...

Moreover, while Dario talks about pacing, the computing bill in his hands shows absolutely no sign of pacing.

Reuters disclosed that, to date, Anthropic has committed to spending more than $100 billion on Amazon cloud services over the next ten years.

There is also a Microsoft Azure partnership worth around $30 billion, likewise backed by large-scale Nvidia GPU computing power.

So A company currently gives people a rather curious impression—

Says: AI capabilities are advancing too quickly; does: Bring on more GPUs.

Says: It would be best for the entire industry to slow down together.

Does: $65 billion in funding received, $100 billion computing agreement signed, and $2 trillion IPO continuing to move forward.

No truth in what it says, yet it is anxious and unable to get motivated—no wonder it hates DeepSeek's open source...

As the old saying goes, nobody likes saying one thing and doing another.

It seems Silicon Valley has collectively become desensitized to the phrase “for your own good”??