Recently, OpenAI has seen another departure, as Miles Brundage, the head of the Safety Policy Advisory and AGI Readiness Team, announced his departure from the company. He also revealed that the team he led will be disbanded, with members being reassigned to other departments.
Brundage posted an article on the social media platform X detailing his decision. He stated that the main reason for leaving OpenAI was his desire for more independence and freedom to publish research.
It is noteworthy that Brundage is not the only senior security researcher to leave. As early as May this year, OpenAI disbanded its Superalignment Team, which was primarily responsible for risk assessment of artificial superintelligence. Since then, the team's two leaders, Jan Leike and Ilya Sutsver, have also left the company.
Additionally, the company's Chief Technology Officer, Mira Murati, Chief Research Officer Bob McGrew, and Vice President of Research Barret Zoph have also left recently.
Over the past six years, Brundage has been advising OpenAI's management and board, helping the company prepare for the rapid development of artificial intelligence. His contributions in safety research are significant, including the introduction of an external red teaming mechanism, which allows external experts to detect potential issues with OpenAI's products.
Brundage mentioned disagreements with OpenAI over research restrictions, feeling that these restrictions were becoming increasingly difficult to accept. He believed that his work within OpenAI had influenced his research objectivity, making it difficult for him to remain impartial on future AI policy issues. He also mentioned on social media that there is a general view within OpenAI that "the cost of speaking out is high, and only some can do this."
This series of executive departures has not only drawn more attention to OpenAI's future direction but also raised more questions about its internal culture and decision-making mechanisms.
Key Points:
🌟 Brundage announced his departure from OpenAI, and the AGI Readiness Team will also be disbanded.
💡 He hopes to have more freedom to publish research, reflecting disagreements with the company over research restrictions.
🚀 Multiple executives have left recently, drawing attention to OpenAI's future development and internal culture.