Former OpenAI's super alignment team lead joins rivals Google-backed Anthropic
1 min. read
Published on
Share this article
Improve this guide
Read our disclosure page to find out how can you help MSPoweruser sustain the editorial team Read more
Key notes
- AI safety researcher Jan Leike joins Google-backed Anthropic after leaving OpenAI.
- Leike’s prior role focused on long-term risks of advanced AI.
- His new research areas include scalable oversight and automated alignment of AI.
Jan Leike, previous head of the superalignment team who recently departed OpenAI, along with a co-founder, has joined competitor Anthropic, an AI startup backed by Google and Amazon.
Leike’s move comes after his resignation from OpenAI earlier this month, following the disbanding of the superalignment group he co-led. This group, formed in 2023, focused on researching long-term risks associated with advanced AI.
Leike announced his new position at Anthropic on social media, showing his enthusiasm for continuing “the superalignment mission.” His new role will involve research on scalable oversight, weak-to-strong generalization, and techniques for automated alignment of AI systems.
Anthropic has secured a huge funding, with Amazon pledging up to $4 billion for a minority stake.
In response to safety concerns, OpenAI, backed by Microsoft, announced the formation of a new safety and security committee led by its CEO and senior executives. This committee will advise the company’s board on safety and security decisions for OpenAI’s projects and operations.
Anthropic, founded in 2021 by former OpenAI executives, launched its own large language model, Claude 3, in March to compete with OpenAI’s ChatGPT.
The company has received funding from various tech giants, including Google, Salesforce, and Zoom, alongside Amazon.
User forum
0 messages