Anthropic Files $965B IPO, Trump Signs AI Executive Order, and ChatGPT Crosses 1B Users | EP #262

Anthropic confidentially filed IPO paperwork with the SEC at an estimated valuation of approximately $965 billion, implying a revenue multiple of roughly 20 times on estimated annual revenue of $50 billion to $60 billion. The company generates approximately $9.4 million in revenue per employee, about four times Apple and Google on that metric, and reached a near-trillion-dollar valuation in roughly five years compared to ten years for OpenAI and 21 years for Google. Polymarket gives a 60 percent chance Anthropic surpasses $1.8 trillion in market cap on its first day of trading, and speakers predicted the IPO will be oversubscribed given approximately one trillion dollars of capital seeking generative AI exposure. Speakers noted that going public forces disclosure, which is appropriate for a safety-conscious company.

Trump signed an executive order rejecting heavy AI regulation and asking labs to voluntarily share models 30 days before public release. About three weeks before signing, Trump received calls from David Sacks, Elon Musk, and lab heads saying the original order was overreaching, and the result was a watered-down version that Sam Altman and Anthropic both endorsed. Dave characterized the outcome as self-regulation achieved through personal meetings with the White House and cautioned that governing through voluntary compliance is not good in the long run. Emad argued that even 30 days of advance model access is a massive tactical advantage for the US government, and Alex noted that governmental functions like discovering zero-day vulnerabilities at scale are now de facto privatized.

ChatGPT passed one billion monthly active users roughly three years after launching, compared to a decade for YouTube and five years for TikTok, with year-over-year growth at 62 percent and holding. Claude sits at 56 million monthly active users but is growing at 640 percent year-on-year, approximately ten times ChatGPT's growth rate. Sam Altman has said he would choose a billion monthly active users over the strongest model, viewing distribution as a stronger moat than model weights. Alex assessed that GPT-5.5 is currently the strongest model available, meaning OpenAI now holds both the broadest distribution and the strongest model simultaneously. Imad predicted the most important agent will be the one coordinating all other agents closest to the user, and that this will cause user acquisition costs to rise sharply as companies pursue IPOs and shift to paid acquisition.

OpenAI launched Rosalind BioDefense, giving trusted researchers and government public health agencies access to AI tools for outbreak detection, disease surveillance, and vaccine development. Alex observed that the G in AGI is starting to shrink as advanced bio and cyber capabilities get carved out into separate models available only to exclusive audiences, and warned that the same technology enabling biodefense can be used by terrorists to develop new viruses. Sam Altman, Dario Amodei, Demis Hassabis, and Mustafa Suleyman co-signed an open letter to Congress calling for laws requiring DNA synthesis companies to screen orders, as dozens of companies currently print custom synthetic DNA without screening customers. Companies like Twist Biosciences already screen and have lobbied for legislation, but screening costs money and creates competitive disadvantage if only responsible players comply, making legislation necessary.

Microsoft announced seven in-house AI models at Build 2026 and a collaboration with Mayo Clinic on a healthcare frontier model, but Alex argued Microsoft is not truly back in the AI game and Imad characterized it as currently at the level of a good Chinese lab unlikely to advance further toward AGI. Imad cited the departure of the Wizard LM team to Tencent as evidence researchers could not pursue AGI-level work at Microsoft, and Dave noted that Mark Chen turned down a billion-dollar compensation package from Meta to leave OpenAI, suggesting Microsoft is unable or unwilling to offer compensation necessary to attract top frontier researchers.

Bernie Sanders introduced the American AI Sovereign Wealth Fund Act, which would impose a one-time 50 percent tax not on profits but on the stock of the largest AI companies, with proceeds funding a public wealth fund. Dave argued the proposal is fundamentally flawed because distributing equity to citizens requires selling the stock, and dumping five trillion dollars of stock on the market would collapse valuations. Imad calculated that half of Anthropic and half of OpenAI combined would amount to only a couple thousand dollars per American citizen, and warned the structure would make these companies too big to fail and entrench whoever controls the fund as an outsized political power.

Nvidia is launching ARM-based PC processors, the N1 and N1X, with the N1X featuring GPU performance on par with the RTX 5070 in a laptop chip, positioning it as a direct challenge to Apple, Intel, and AMD. Nvidia currently has no direct consumer contact or consumer data, making the PC chip launch a strategic move into consumer-facing markets. Nvidia now represents the majority of transistors produced at TSMC's most advanced node, surpassing Apple, and is approximately eight to ten times Intel's market value. Nvidia is pursuing an open source strategy through its Cosmos and Nematron alliances to own the home AI market and block AMD at the edge.

New Limit, co-founded by Brian Armstrong, raised $435 million at a $3.1 billion valuation and is developing epigenetic reprogramming therapies to reverse cellular aging, with human clinical studies targeted for next year.