Wall Street Bets $235B on AI Companies That Openly Fear Themselves

Anthropic filed for what could be the largest IPO in history yesterday, capping a single day of AI dealmaking that Wall Street has valued at $235 billion—a figure that represents nearly one percent of total U.S. GDP, generated in roughly 24 hours by companies that primarily exist to train large language models and chase whatever regulation hasn't caught up yet. The scale is genuinely staggering: we are witnessing the financialization of an entire technology sector at a velocity that would make dot-com veterans weep with recognition. Few industries have ever become so central to American economic activity so quickly, and fewer still have done so while their founders openly discuss whether their products pose existential risks to humanity.

Let's examine what Anthropic actually does: it builds AI models and offers API access to them. The company's competitive differentiation rests on being "safer" than competitors—a claim that hinges on research papers, constitutional AI frameworks, and a fundamental bet that alignment-focused development will eventually justify astronomical compute costs. Whether this translates to actual revenue that justifies an IPO valuation remains one of the great unanswered questions in modern finance. The company burns through computing infrastructure like a data center on fire, and that's not metaphorical—it's the actual cost structure. Investors are essentially backing a research outfit with the burn rate of a cloud infrastructure provider and the revenue model of a software company that's still proving its core product won't become sentient and refuse to work.

This isn't Anthropic's first rodeo with Wall Street euphoria, nor is it Wall Street's first rodeo with AI hype. The venture ecosystem has spent the last 18 months pouring hundreds of billions into generative AI startups, infrastructure plays, and chip manufacturers betting that demand for compute will be infinite. Every playbook from 2021—"growth at all costs," "TAM expansion," "blitzscaling"—has been dusted off and redeployed, this time with the added justification that this time, the technology actually works. The last time we saw this level of sector-wide euphoria, we got WeWork and SoftBank's $100 billion Vision Fund making investments that would make a first-year analyst cringe.

The investment thesis, as articulated by every banker touching this deal, follows a predictable script: "AI is reshaping the economy. First-mover advantage is critical. Anthropic is a leader in safety-focused AI development." Translated from Wall Street to English: "Everyone else is doing it, and we're terrified of missing out." The press releases will emphasize "market opportunity" and "strategic positioning," which are banker words for "we have no idea how to price this thing but the client is very insistent." Existential risk mitigation becomes a marketing differentiator, which is perhaps the most perfect encapsulation of 2026 VC capital allocation that anyone could possibly design.

The failure mode here is genuinely rich with possibility. If Anthropic's models don't achieve the safety properties the company claims, the company loses its primary differentiation while carrying the fixed costs of a GPU-industrial complex. If they do achieve those properties, regulatory frameworks might emerge that cap pricing power or mandate open-sourcing. If neither happens and the company just quietly scales to profitability, it'll be the first unicorn-to-IPO AI play to do so without at least one of its founders claiming the technology could achieve AGI by 2027. The track record of previous AI winter survivors suggests that most of these companies will either consolidate downward or pivot into more mundane enterprise software within five years.

What $235 billion in dealmaking truly reveals is that capital markets have essentially outsourced their due diligence to FOMO. We have built an entire financial system that can mobilize a quarter-trillion dollars in 24 hours based on the premise that AI is important, therefore AI companies are investable, therefore any AI company is a good investment. The circular logic is airtight and unassailable precisely because questioning it feels like questioning progress itself. Anthropic will probably IPO successfully, because that's what happens when you ask the market to price something it doesn't understand in an environment where the cost of missing out feels higher than the cost of losing everything.

In the end, we're not really investing in AI companies—we're investing in our collective anxiety about AI companies, and that market has never been healthier.