Via a brief news release on Oct. 26, Anthropic said that it had filed a registration statement on Form S-1 with the U.S. Securities and Exchange Commission (SEC) regarding its upcoming initial public offering (IPO). There's not much detail on the release, but the prospectus with the SEC shows that the San Francisco-based company that developed the AI chatbot Claude sees a five-year power bottleneck. The company is moving from a cloud-only, fixed-time compute model to a hybrid setup that includes directly leased chips hosted in data center facilities. In the prospectus, Anthropic discusses launching a new initiative, "Anthropic Direct," to build its own dedicated data centers rather than relying entirely on third-party cloud providers. The company expects this move to lower its long-term costs. Partner companies will buy the land, secure power, build the facilities, and run the daily operations. Anthropic will then lease the computing power directly from those partners. Last month, Anthropic began signing these agreements, committing to $125.2 billion in lease payments over five years, before interest.
CryptoAlerta — análise de criptomoedas e mercado em tempo real