There’s lots of news circulating about the big AI labs at the moment, but the smaller labs that are working in stealth seem to be recruiting some top talent of their own.
Former OpenAI cofounder John Schulman has left Anthropic and is going to join former OpenAI CTO Mira Murati’s new startup, Fortune reports. Schulman had left OpenAI to join Anthropic just five months ago. He now seems set to join the startup founded by former OpenAI CTO Mira Murati, who had also left OpenAI around five months ago to start her own company.
“Confirming that I left Anthropic last week,” Schulman had posted on X. “Leaving wasn’t easy because I enjoyed the stimulating research environment and the kind and talented people I was working with, but I decided to go with another opportunity that I found extremely compelling. I’ll share more details in the coming weeks. Thanks to Jared, Jan, Dario and others for the support during my time at Anthropic, and I wish them all the best,” he had added.
John Schulman had joined OpenAI shortly after earning his Ph.D. in computer science from the University of California, Berkeley, in 2016. At OpenAI, he led the reinforcement training org, which fine-tunes generative AI models to follow human instructions, and later became the head of OpenAI’s alignment science efforts. He’d left OpenAI to join Anthropic last year, and has now left Anthropic to join Mira Murati’s new startup.
Not much is known about Mira Murati’s new company. It doesn’t even have a publicly declared name or mission yet, but already seems to have some former OpenAI researchers working for it. These reportedly include Christian Gibson, who had been serving on OpenAI’s supercomputing team, Barret Zoph, who had left OpenAI around the same time as her, and Jonathan Lachman, who had led special projects at OpenAI.
Mira Murati isn’t the only former OpenAI executive that has launched their own company. Anthropic, which makes the Claude series of models, was also founded by former OpenAI researchers. Most prominently, former OpenAI Chief Scientist Ilya Sutskever had left the company to launch his own startup, named SSI, or Safe Superintelligence. SSI has given out no updates at all since being founded, and is believed to have a single-shot aim of producing superintelligence. And this dispersion of talent across several labs might be safe in itself — it increases the probability of ensuring that no single company will create or control AGI when it finally does come to pass.