When Unicorns Become Common, a New Superanimal Is Born: The Leviathan

There was a time when a billion-dollar startup sounded impossible. In 2013, venture capitalist Aileen Lee gave those companies a name: unicorns. They were rare, almost mythical, creatures of Silicon Valley imagination. A private startup valued at more than $1 billion was not just a company; it was a sighting. Investors hunted for them. Founders dreamed of becoming one. The word worked because it captured scarcity.

But the age of artificial intelligence has done to valuation what inflation does to money: it has changed the emotional value of the number. A billion no longer sounds like a miracle. In the AI boom, it can sound like a seed round with ambition. Unicorns have not disappeared; they have multiplied until the metaphor itself has weakened. A billion-dollar valuation, once the horn of a mythical animal, now resembles a wild horse on the venture-capital plains.

That is why we need a new creature. Above the unicorn, above the decacorn, above the hectocorn, there is another scale entirely: the trillion-dollar startup, or near-trillion-dollar startup, so vast that it no longer behaves like an ordinary company. It begins to resemble infrastructure, government, weather, mythology. For that creature, the right name is Leviathan.

The Leviathan is not a cute mascot. It is a biblical sea monster, a symbol of overwhelming scale and chaos; in later political thought, especially through Thomas Hobbes, the word also came to evoke the giant institution, the sovereign body, the artificial person bigger than any individual inside it. That double meaning makes it perfect for frontier AI. A Leviathan is not just valuable. It is so large that it becomes a force of artificial nature.

Anthropic, the company behind Claude, is now approaching that category. On June 1, 2026, Anthropic announced that it had confidentially submitted a draft S-1 registration statement to the U.S. Securities and Exchange Commission for a proposed IPO. The company stressed that the number of shares and the price had not yet been set, and that any listing would depend on SEC review, market conditions and other factors. A few days earlier, it had announced a $65 billion Series H round at a $965 billion post-money valuation.

That is not yet a trillion-dollar IPO. It is something more suspenseful: a near-trillion-dollar private company walking toward the public markets, still partly underwater, like a sea monster before it breaks the surface.

At the center of that creature stands Dario Amodei, Anthropic’s co-founder and CEO. His biography is not the usual mythology of the college dropout who saw the future before professors did. It is almost the opposite. Amodei is a product of deep formal education: Lowell High School in San Francisco, Caltech, Stanford physics, Princeton graduate work, neural circuits, biophysics, electrophysiology, and a postdoctoral appointment at Stanford Medicine. Wikipedia records that he was born in San Francisco in 1983; his father, Riccardo Amodei, was an Italian-American leather craftsman from Tuscany, and his mother, Elena Engel, was a Jewish American from Chicago who worked as a project manager for libraries.

That background matters, but not because ancestry explains genius. It matters because it places Amodei in a lineage of craft and scholarship: a father who worked with material things, a mother connected to libraries, and a son who moved through physics into the study of neural systems before building machines that imitate, extend and disturb human cognition. His path is a reminder that the AI revolution did not emerge from nowhere. It was built by people trained in hard disciplines, in laboratories, in mathematics, in the habits of long attention.

This is especially ironic because AI is often discussed as a replacement for education. The story goes that students will no longer need to learn writing, coding, mathematics or even reasoning, because machines will do the work for them. Amodei’s life argues the opposite. The people building the most powerful AI systems are not proof that education is obsolete. They are proof that education compounds. His doctoral work at Princeton focused on neural circuits; the Hertz Foundation describes his research as involving statistical mechanics models of neural circuits and tools for neural recording. Before Anthropic, he worked on machine learning in elite research environments, including Google Brain and OpenAI.

The lesson is not that everyone needs a Princeton PhD. The lesson is that AI makes shallow knowledge cheaper and deep knowledge more valuable. When machines can generate fluent answers, the human advantage shifts toward judgment: knowing what to ask, what to distrust, what to measure, what to build, and what not to release. Education is not merely the transfer of facts. It is the formation of taste, skepticism, discipline and responsibility. In the age of Leviathans, that formation becomes more important, not less.

Amodei’s professional path also runs through the center of the modern AI story. Before founding Anthropic, he was a senior research scientist at Google Brain. He then joined OpenAI, where he became vice president of research and helped lead work on large language models including GPT-2 and GPT-3. Those models were part of the technical prehistory of ChatGPT, the product that made generative AI a household phenomenon.

Then came the split. In 2021, Dario and his sister Daniela Amodei co-founded Anthropic with other former OpenAI employees. The difference was not simply commercial rivalry. Anthropic presented itself as an AI safety and research company, focused on systems that are steerable, interpretable and reliable. In other words: it was not enough to build the monster; one also had to understand how to restrain it.

The rivalry with OpenAI became almost Shakespearean in November 2023, when OpenAI’s board briefly removed Sam Altman as CEO. During that crisis, Reuters reported that the OpenAI board approached Amodei about replacing Altman and potentially merging OpenAI and Anthropic. Amodei declined both the leadership offer and the merger proposal.

That refusal now looks historically important. Had Amodei accepted, the AI world might have consolidated around a single dominant institution. Instead, Anthropic remained separate, and the industry kept two competing poles: OpenAI, the maker of ChatGPT, and Anthropic, the maker of Claude. Today, their rivalry is no longer just about chatbots. It is about corporate structure, safety philosophy, enterprise customers, capital markets and the future governance of machine intelligence.

Anthropic’s rise has been powered in large part by enterprise adoption. The company says Claude is being deployed across industries and that its revenue run rate crossed $47 billion in May 2026. Its products include Claude, Claude Code and Claude Cowork, with coding and professional work becoming central to its business case. The Dutch article’s point is right: Anthropic is not selling a futuristic fantasy alone; it is selling tools that companies already use inside workflows.

The more unsettling symbol is Mythos. Anthropic’s Project Glasswing announcement describes Claude Mythos Preview as an unreleased frontier model whose cyber capabilities have already found thousands of high-severity vulnerabilities, including in major operating systems and browsers. Anthropic frames this as a defensive initiative, but the implication is obvious: a system that can find and exploit software flaws is valuable to defenders and dangerous in the wrong hands. The U.K. AI Security Institute reported that Mythos Preview showed major progress on capture-the-flag tasks and multi-step cyber-attack simulations.

Here the Leviathan metaphor becomes more than branding. A unicorn is charming. A Leviathan is ambiguous. It may protect the harbor, or it may sink ships. It may secure the world’s software, or it may reveal how fragile that software always was. It may become a tool of productivity, medicine, science and education, or it may become the infrastructure through which power concentrates faster than democratic institutions can react.

That is why Anthropic’s possible IPO matters beyond Wall Street. For early employees and private investors, it could be a payday of historic proportions. For public investors, it could be the first chance to buy into one of the frontier AI labs directly. For index funds and ETFs, a near-trillion-dollar public Anthropic could eventually become impossible to ignore. Reuters noted that a company of this scale would sit among the top tier of the S&P 500 and could reshape benchmark indexes and investor flows.

But the public markets will also force a new kind of visibility. Private valuations can be sustained by narratives, scarcity and elite confidence. Public companies must report. They must disclose. They must explain margins, capital expenditure, revenue quality, customer concentration, risks and losses. The IPO is therefore not merely Anthropic’s concert premiere, as the Dutch article nicely suggests. It is also the moment when the fog begins to lift from the monster’s outline.

The skeptics are not automatically wrong. Every technological boom produces two things at once: real transformation and financial hallucination. The dot-com era gave the world Amazon, but it also gave it wreckage. AI may do the same. A trillion-dollar valuation can be a prophecy, but it can also be a fever. Michael Burry and other doubters look at these numbers and see a bubble: too much capital, too much expectation, too little proof that the economics can justify the myth.

Yet even if the valuation falls, the creature has already been born. The deeper issue is not whether Anthropic is worth $965 billion, $500 billion or more than $1 trillion. The deeper issue is that AI companies are beginning to occupy a scale once reserved for states, central banks, energy systems and global infrastructure. They are not merely startups. They are private institutions building tools that may reorganize labor, education, cybersecurity, medicine, warfare and knowledge itself.

That is what makes the word Leviathan useful. It captures both size and uncertainty. It says: this is bigger than a unicorn. It is not just rare; it is world-shaping. It is not merely desirable; it is dangerous. It is not a horse with a horn; it is a sea monster beneath the surface of everyday life.

Dario Amodei’s personal story gives the monster a human face: the son of an Italian-American craftsman and a Jewish-American library project manager, a physics student, a biophysicist, a Google Brain researcher, an OpenAI leader, and finally the CEO of Anthropic. But the company he built now exceeds biography. It belongs to a new category of institution.

When unicorns become common, we should stop pretending that the old myths are enough. The AI age has inflated the billion into something ordinary. The next creature has arrived, and it is not gentle. It is the Leviathan: vast, intelligent, useful, frightening, and still only partly visible above the water.