TL;DR
The company behind the Claude AI models has told would-be investors that its next systems "may imminently" reach a point it has been watching for years: AI that speeds up the making of AI, to roughly twice the old pace, or two years of progress packed into one. That line comes from Anthropic's amended IPO filing, as quoted in a published report, and it is a risk disclosure rather than a test result; the company says no current model is rated past the line. If it arrives, the payoff could reach well beyond chatbots. Faster AI research is the kind of engine that could pull the dates for new drugs, new materials and new science earlier across the board, and the company says it intends to keep humans in charge of the process. Our odds that Anthropic rates a model past this threshold: chance it works 85%, likely by 2027. To see the full breakdown of the odds and the timeline of breakthroughs that leads there, subscribe.
Below the line, for paid subscribers: The full story: where it stands, how we got here and what's dated next; our labeled speculation on what it adds up to; and every source.