The AI lab revolving door spins ever faster

In Brief

AI labs conscionable can’t get their employees to enactment put.

Yesterday’s large AI quality was the abrupt and seemingly acrimonious departure of 3 apical executives astatine Mira Murati’s Thinking Machines lab. All 3 were rapidly snapped up by OpenAI, and present it seems they won’t beryllium the past to leave. Alex Heath is reporting that 2 much employees are expected to permission for OpenAI successful the adjacent fewer weeks.

Meanwhile, Anthropic continues to propulsion alignment researchers distant from OpenAI. The Verge is reporting that 1 of OpenAI’s elder information probe leads, Andrea Vallone, has near the institution for Anthropic. Vallone specializes successful however AI models respond to intelligence wellness issues — which is simply a peculiarly delicate contented for OpenAI aft its caller sycophancy problems. As The Verge notes, Vallone volition beryllium moving nether alignment researcher Jan Leike, who near OpenAI successful 2024 implicit concerns the institution wasn’t taking information earnestly enough.

If that wasn’t enough, OpenAI finished things disconnected with 1 past large poach. Max Stoiber, formerly the manager of engineering astatine Shopify, volition beryllium joining the company to enactment connected OpenAI’s long-rumored operating system, successful what helium describes arsenic a “small high-agency team.”

Subscribe for the industry’s biggest tech news