OpenAI key researcher Jan Leike, who resigned earlier this month citing “security concerns,” has joined rival Anthropic to “work on scalable supervision, weak-to-strong generalization, and automatic alignment,” according to Leike’s post on X Research”.
Anthropic positions itself as the antithesis of OpenAI: a public benefit company dedicated to cautious progress. According to the company, it was founded by former OpenAI employees who left the company. Financial TimesBecause they believe OpenAI has strayed from its mission of developing safer and more ethical artificial intelligence and has become too focused on commercial interests after its $1 billion deal with Microsoft. For those who follow artificial intelligence TV series, Lake joining Anthropic will come as no surprise.
“It’s long overdue that we take the impacts of AGI seriously. We must prioritize being as prepared for them as possible,” Lake said in a follow-up post about his resignation a few weeks ago. “Only in this way can we ensure that AGI benefits all mankind.”
Leike previously worked on OpenAI’s “Superalignment” team, which seems consistent with his role at Anthropic, ensuring that intelligent systems are designed to align with human values. In a post after announcing his resignation, Lake criticized OpenAI for not providing enough computing resources to ensure success.
For its part, Anthropic uses “constitutional AI” to incorporate ethical principles into the training of its AI models by gathering public input and integrating concepts from sources such as the United Nations Universal Declaration of Human Rights, as I’ve reported previously .
Another notable departure from OpenAI is that of its co-founder Ilya Sutskever, who departed earlier this month after reportedly leading a campaign against CEO Sam Sam in November. He has been absent from office since Sam Altman’s failed coup. Although Sutskever resigned and has yet to indicate his next move, his values align with Anthropic, making it a logical destination for him alongside Leike.
