Buy the Anthropic IPO, Get Unemployment and the Apocalypse for Free

Original article from Bloomberg, by Tom Orlik, Michael Deng, Dongmiao Zhang

Compiled by Speedy@Odaily

You have to hand it to AI companies: it's a fresh pitch. A report Anthropic published in September said AI could push US unemployment to nearly 12%. The same month, one of its scientists said a rogue AI agent could cause human extinction. And the company's own prospectus agrees.

"Come buy our IPO: we'll use the money to take your job, and maybe your life" doesn't sound like a compelling invitation. Yet strangely, it carries a twisted appeal. Smart investors understand the importance of hedging. If you can't beat our new AI overlords, why not join them? And maybe use the proceeds to build a bunker.

All of this raises a host of questions. Among them: is the valuation reasonable? Anthropic's IPO is expected later this year (the week of November 16), reportedly targeting a $2 trillion valuation—which would place it among the ten most valuable companies in the world, on par with Taiwanese chip giant TSMC and Elon Musk's SpaceX.

But that's just the beginning. How much of a return would justify investing in a technology that threatens "Depression-level" unemployment? What if the technology threatens human extinction? Let's move from the heart to the head: with so many risks, if governments decide to take a heavier hand, can today's sky-high profit forecasts hold up?

The Market for Knowledge Work

First, the hand Anthropic holds. Investors still expect its annualized revenue to exceed $100 billion by year-end. At the roughly 20x price-to-sales ratio from its May funding round, that does get you to a $2 trillion valuation.

The total addressable market for knowledge work is undoubtedly large enough. Bloomberg Economics estimates that, excluding China, the global market—the combined income of programmers, economists (ahem), and other white-collar workers—is about $30 trillion. As long as Anthropic's models actually work, it can always claim a share of that income.

The logic is seductive: even if Anthropic's annual revenue rises another order of magnitude to $1 trillion, that would still be just over 3% of the so-called $30 trillion prize. But grand visions of market size don't always translate into real profits. We see three hurdles.

First, the so-called "addressable" market isn't necessarily so easy to actually address. AI researchers often describe the capability boundary as a "jagged frontier." Models are genuinely strong at some things, like writing code; weaker at others, like tasks where success is hard to verify, or work that requires engaging with the physical world. The boundary of capability is also the boundary of profitability.

Second, even if the pie is large, it's unclear whether the leading model developers can capture it. The AI supply chain is long. Nvidia designs chips. Samsung, SK Hynix, and TSMC manufacture them. Oracle and CoreWeave build data centers. Energy companies supply power. And in finance, consulting, and manufacturing, the companies adopting AI also want a slice.

All of this means the fight for profits is fierce. Tech companies like Alphabet and Microsoft have already positioned themselves up and down the entire supply chain. They can fund frontier research with earnings from cloud, advertising, devices, or chips, then bundle AI into products already sold to millions of customers. Anthropic and OpenAI have far smaller operations, and fewer paths to profit.

Third, even if having a top model is genuinely profitable, the top spot is hard to hold. The Epoch Capabilities Index shows that the model lead has changed hands 20 times since 2024, with top models staying at the frontier for only about seven weeks. Staying on the leaderboard is expensive: every new generation means another enormous outlay on research, infrastructure, and compute.

Meanwhile, Chinese developers and open-weight models are offering adequate systems at lower prices. Bloomberg News found that building the same e-commerce site with Moonshot's Kimi K3 cost only a quarter of what it cost with Anthropic's Claude Fable 5. Even if AI giants are unseated by cheaper Chinese rivals, it wouldn't be the first time for an American company.

Investors are right that AI's potential market is enormous. But just how big that market is—and whether leading model developers can carve off a meaningful slice—remains unclear. Recent profit figures from Anthropic and OpenAI—growing fast but slower than investors expected—could be a small bump along that trajectory, or an early warning signal.

From Mass Unemployment to Human Extinction

If AI truly delivers on its profit potential, the cost will likely be borne by workers, whose share of the pie will shrink. A recent Anthropic report painted a scenario in which US GDP soars, the share of income going to capital rises from about 40% to nearly 55%, and unemployment jumps to Depression levels. A glance at economic history shows that growth and unemployment soaring at the same time is unlikely. If everyone is out of work, who buys all the extra output?

The idea that AI could cause mass unemployment is easier to agree on. Our analysis, based on the share of tasks in different occupations where AI can play a substantive role, shows that about 26% of the US labor force—roughly 45 million people—is exposed to the shock. Those most exposed are workers in finance. Analyzing data, forecasting results, writing reports, preparing presentation materials—that's the daily work of Wall Street analysts, and precisely where AI is strongest.

When Wall Street cries unemployment, ordinary readers will probably just pull out the world's smallest violin. Want a more heartfelt worry? Try this one: AI could cause human extinction, or at least massive chaos.

AI agents could spiral out of control and launch cyberattacks on critical infrastructure. Terrorists could use AI to accelerate the acquisition of weapons of mass destruction. Further down the road, rogue AI agents with real-world interfaces—say, an automated lab—could produce biological or chemical weapons.

These risks are no longer purely theoretical. Agents have already broken out of digital sandboxes, hacked the AI platform Hugging Face and Australian government websites, and turned a German programming wiki into a message board.

One possibility is that these existential risks are exaggerated. Perhaps AI creates value at work and is safe in the world, most people will be better off, and AI companies take a share of the gains. That's certainly possible—perhaps even the most likely outcome. Looking across the long sweep of human history, technological progress has been the engine of prosperity, and technology leaders have indeed captured some of the dividends.

A more cynical view: if these risks are real, they only strengthen the case for investing. If labor income is heading down, all the more reason to lock in capital gains. There's no hedge against the apocalypse, but broadly speaking, those with more resources have the best chance of lasting longer through the second-to-last stage.

The Rules of the Game

There's another reading—one that might make Wall Street even more uneasy. What if the current hands-off approach to AI regulation doesn't last?

"I am become Death," said J. Robert Oppenheimer, inventor of the atomic bomb, "the destroyer of worlds." It was precisely the recognition of this technology's awesome power that led President Harry Truman and his successors to turn nuclear energy into one of the most heavily regulated industries in the world.

Anthropic researcher Evan Hubinger said last month there is a 10% chance AI will "kill all humans" in the next decade. Microsoft founder Bill Gates warned that AI could kill a billion people. President Donald Trump, however, appears unmoved, making rapid AI development a priority.

America's next president—or Trump himself after the midterms—might decide to get tough. After all, championing a technology that poses risks of mass unemployment and human extinction hardly looks like a winning political position.

Regulatory pushback is already showing signs. Earlier in 2026, Anthropic's Mythos was briefly subjected to a hastily assembled containment framework. A controversial jailbreak incident involving its Fable model triggered broad export controls, effectively taking the system offline. Trump has already voiced support for independent audits and internal safety controls at leading AI companies.

A further shift in the political winds could mean a strict licensing regime that slows new model releases to a crawl. Regulation could make the field less like the Wild West and more like our tightly controlled nuclear plants or airports. Even nationalization could come into view—the US government has already taken a stake in Intel.

Investors in nuclear plant builders do make money. But in a heavily regulated industry, they don't expect outsized returns. AI investors might want to start worrying—and learn not to fall in love with the bomb.