Anthropic has reset AI expectations by hiring former OpenAI co-founder Andrej Karpathy, achieving its first profitable quarter, and deepening a compute partnership with SpaceX. The episode covers OpenAI's impending IPO, a potential new AI executive order, and the accelerating race toward recursive self-improvement (RSI) in AI research.
Summarized by Podsumo
Andrej Karpathy, a founding member of OpenAI, has joined Anthropic to lead a team focused on using Claude for recursive pre-training research, signaling a move toward recursive self-improvement (RSI).
Anthropic reported its first profitable quarter ever among foundation labs, forecasting $10.9 billion in Q2 revenue and a small operating profit of $559 million, far ahead of earlier profitability timelines.
OpenAI is filing for an IPO as soon as Friday, with a target to be ready by September, while Anthropic is reportedly targeting October—changing the competition for public market liquidity.
SpaceX has secured a massive $45 billion three-year deal with Anthropic for compute capacity at the Colossus data centers, making it SpaceX's largest revenue source and underscoring the value of compute.
Nvidia's earnings revealed record data center revenue growth of 92%, with CEO Jensen Huang emphasizing that AI infrastructure expansion is accelerating faster than any prior technological buildout.
"I think the next few years at the frontier of alignment will be especially formative. I'm very excited to join the team here and get back to R&D."
"Andrej Karpathy"
"This is like a leading franchise recruiting someone who simultaneously the best player and the league's best broadcaster and its most watched developmental coach all in one."
"Signal"
"The Anthropic IPO won't be for less than $2 trillion."
"Conor Sen"