Tech Executive Confirms Yesterday’s AI Breakthrough Is What He Has Been Saying All Along

When Anthropic published a highly technical paper detailing a fundamentally new approach to continuous reinforcement learning, CTO Noah Tall was already nodding along. He opened X, skimmed a summary thread written by the lead researcher, and immediately quote-posted the thread with his own commentary.“I have been saying this for months,” Tall wrote. “I’ve said this since at least last year.”Tall’s mastery of web development has recently expanded to encompass his new favorite domain: being one of the world’s leading authorities on frontier machine learning. He does not conduct research or build anything himself anymore, but his intuition has nevertheless guided the entire field, often without the field’s knowledge.“I laid this exact architecture out at our last sprint retro,” Tall messaged the company’s “AI Futurists” Slack channel. “I advocated for this exact tensor logic during the engineering all-hands in February. This was part of my undergraduate thesis before neural nets became…