Former OpenAI Safety Lead Joins Anthropic
Less than two weeks after leaving OpenAI, Jan Leike, one of the company’s former artificial intelligence safety leads, announced that he has landed at one of its rivals, Anthropic.
Leike’s New Role at Anthropic
Leike, who co-led OpenAI’s now-disbanded “Superalignment team,” which focused on AI’s existential dangers, resigned from the company earlier this month. He expressed his excitement to join Anthropic and continue the superalignment mission. Leike mentioned that his new team at Anthropic will focus on scalable oversight, weak-to-strong generalization, and automated alignment research.
Anthropic’s Recent Achievements
Anthropic, which recently received a $4 billion investment from Amazon, unveiled its latest AI model family, Claude 3, in March. This new model outperformed OpenAI’s GPT-4. Founded in 2021 by former OpenAI employees Daniela and Dario Amodei, Anthropic aims to develop “safer” AI technologies. The Amodei siblings praised their Claude 3 Opus model as the most capable AI model globally during Bloomberg’s Technology Summit.
Challenges at OpenAI
Leike’s departure from OpenAI came after his co-lead, Ilya Sutskever, announced his own exit. Leike expressed his disappointment with OpenAI’s leadership, stating that the company had not prioritized the safety aspects crucial for the development of artificial general intelligence (AGI). The Superalignment team, despite its vital mission, did not receive the necessary support from OpenAI, including the computing power it was promised.
Source: Fortune
Image/Photo credit: source url