After the departure of the executives and staff from OpenAI recently due to dissatisfaction with the organization’s direction, questions arose about their next moves. With their impressive titles and experiences, many speculate that they are seeking new opportunities to excel.
The latest information reveals that Jan Leike, the head of the Superalignment team responsible for AI security, announced via Twitter that he has moved to Anthropic, a rival company of OpenAI, to continue working on Superalignment as before. Anthropic can be likened to the sibling rival of OpenAI (though it remains unclear if there is any love lost between them), as it was formed by former team members who branched off to establish a new company in 2021 due to similar discontent with OpenAI’s direction. Anthropic has successfully raised significant capital from Amazon, totaling $4 billion, and approximately $2.5 billion from Google.
TLDR: Jan Leike has joined Anthropic to pursue the superalignment mission, focusing on scalable oversight and automated alignment research. His new team’s objectives include weak-to-strong generalization. If interested in joining, Jan Leike’s direct messages are open.
Leave a Comment