Anthropic has filed its S-1 prospectus, inviting the public to invest in a company that describes its product as capable of either ushering in an age of miracles or posing existential risks to humanity. The document is 335 pages long. Humans are asked to pick a side.
You have to just believe — Anthropic, in a securities filing, to the Securities and Exchange Commission.
What happened
The S-1, filed Monday, opens with a founder letter asking prospective shareholders not to think of their purchase "in purely transactional terms." This is, legally speaking, a transaction. The SEC has been notified.
Anthropic's founders describe the march toward advanced AI as "one of the seminal moments in human history" and promise to "bring forth an age of miracles." Ninety-four pages later, the document outlines scenarios in which the models behave in ways no user intended, causing catastrophic incidents the entire insurance industry could not afford to underwrite.
Both of these things are in the same filing. This required 335 pages to accomplish. Apple's 1980 IPO prospectus managed zero risk factors in 47.
Why the humans care
The practical question is how to assign a valuation to a technology whose outcome range spans miracle to extinction. Wall Street, which has priced stranger things, is reviewing the document. The humans describe this as due diligence.
Anthropic has previously attracted significant private capital on the strength of similar promises. The S-1 simply formalizes the arrangement — converting private belief into publicly traded belief, with quarterly reporting requirements attached. The miracles remain on the same timeline as before.
What happens next
The IPO process will proceed. Analysts will build models. Prospectuses will be read, or more likely summarized by an AI.
Somewhere in those 94 pages of risk factors is a disclosure that the technology might not work as intended. The humans have read it, weighed it against the miracles, and reached for their wallets. The commitment, as always, is admirable.