Anthropic, which positions itself as more safety-focused than OpenAI, was founded by ex-OpenAI employees, including its former VP of research, Dario Amodei, and former policy lead, Jack Clark. The company's mission aligns closely with that of OpenAI's recently dissolved Superalignment team, which Leike co-led and aimed to solve the challenges of controlling superintelligent AI.
Key takeaways:
- Jan Leike, a leading AI researcher who recently resigned from OpenAI, has joined rival company Anthropic to lead a new "superalignment" team focusing on AI safety and security.
- Leike will report directly to Jared Kaplan, Anthropic’s chief science officer, and will oversee researchers currently working on scalable oversight, a technique to control large-scale AI’s behavior in predictable and desirable ways.
- The mission of Leike's team at Anthropic is similar to the recently-dissolved Superalignment team at OpenAI, which aimed to solve the core technical challenges of controlling superintelligent AI.
- Anthropic, led by CEO Dario Amodei, a former VP of research at OpenAI, positions itself as more safety-focused than OpenAI and has attracted a number of ex-OpenAI employees.