Mark Zuckerberg has secured Yang Song, a prominent researcher from OpenAI, to be the new research principal of Meta Superintelligence Labs (MSL). Previously, Song led the strategic explorations team at OpenAI and has begun his new role reporting to Shengjia Zhao, also an OpenAI alumnus who has managed MSL since it launched in July.
This hire follows a recent push by Zuckerberg to recruit top AI talent. He has already brought in at least eleven researchers from leading organizations like OpenAI, Google, and Anthropic. Song’s expertise in processing complex datasets and his contributions to OpenAI’s DALL-E 2 model make him a significant addition to Meta’s team.
Song started with OpenAI in 2022, where his research focused on enhancing models’ abilities to handle diverse data. He holds an impressive academic background, having developed key methodologies during his time as a graduate student at Stanford, where he and Zhao were both mentored by Stefano Ermon.
This staffing change occurred amidst a broader context of shake-ups within Meta’s AI division. Although Zhao was perceived to be replacing Yann LeCun, Meta’s longstanding chief AI scientist, LeCun clarified that he continues in his role at Facebook AI Research (FAIR). Reports indicate a few researchers have already left MSL since its inception, including two who returned to OpenAI and another who joined Microsoft AI after a brief stint at Meta.
Yang Song’s arrival adds to Meta’s ongoing effort to strengthen its position in the competitive AI landscape, contributing to a complex and evolving structure within the company’s AI division.
