Meta AI researcher Andrew Tulloch is leaving the company less than a year after returning from Mira Murati’s Thinking Machines Lab, according to the information cited in the source report. Tulloch had been working inside TBD Lab, the group led by Alexandr Wang and described in the report as a core team in Mark Zuckerberg’s current AI strategy. Semafor said Tulloch had been preparing to leave for some time and waited until Meta rolled out a new batch of open-source AI models and its personal agent product, Muse, before making his exit. The report also said Meta offered him a compensation package in 2025 worth as much as $1.5 billion over at least six years, but he declined at the time. That figure included a maximum bonus component and potential gains tied to a sharp rise in Meta’s share price. Months later, Tulloch still joined Meta. Axios said his eventual pay package was lower than that earlier offer, though the exact amount was not disclosed. Tulloch had previously spent 11 years at Meta, moved to OpenAI in 2023, and later helped launch Thinking Machines Lab. His next destination and the reason for his departure have not been made public.

Jiahui Yu, a core researcher in Meta’s superintelligence effort, has announced that he is leaving the company to start a new venture. Yu joined Meta in late June 2025 and disclosed his departure on August 14, 2026, ending a stint of a little over one year. His personal website has already been updated to reflect the change. The move comes only days after Meta released Muse Spark 1.2, the latest update in a product line developed by the team Yu helped build. In his public statement, Yu said working with CEO Mark Zuckerberg and Meta Chief AI Officer Alexandr Wang to build TBD Lab had been 「inspiring and fulfilling」. He also referenced a question he described as highly important to humanity yet still underexplored, saying it would now receive his full attention. He did not disclose the name of the new company, its research agenda, or any co-founders, adding only that there would be more to share as the work takes shape. Yu was one of the most watched names in Meta’s AI hiring push, during which reports and public comments fueled speculation around compensation packages at the $100 million level. While Meta executives later said such figures did not apply broadly and could reflect stock, bonuses, tenure, and performance conditions, Yu remained closely associated with that recruitment wave. Before Meta, he worked at Google Brain and OpenAI and contributed to projects including Gemini, GPT-4o, and GPT-4.1.

Meta Superintelligence Labs has lost another key researcher. Jiahui Yu, a multimodal AI researcher who joined Meta from OpenAI in 2025, said on Aug. 14 that he is leaving the company to start a new venture. He did not disclose the startup’s name or its exact research agenda, saying only that he plans to work on a question that is “crucial to humanity’s future” yet remains largely unexplored. Yu said on X that he was leaving Meta after working with Mark Zuckerberg and Meta Superintelligence Labs head Alexandr Wang on TBD Lab. He also highlighted the multimodal team’s work on Muse Spark, Voice Mode, Muse Image and Muse Video. Before Meta, Yu led research for OpenAI’s Perception team and worked on GPT-4o, GPT-4o Image Generation, GPT-4.1, o3 and o4-mini. Earlier, he was involved in Gemini multimodal model development at Google DeepMind, where his personal website lists him as a co-lead of Gemini Multimodal. His departure comes after Meta’s 2025 recruiting push into OpenAI, when Sam Altman said some offers included signing bonuses of as much as $100 million.

Meta abandons open-source tradition with closed-source AI model 'Avocado', integrating Google Gemma and Alibaba Qwen. The project led by TBD Lab expects 2026 spring release. Meta reallocates $60 billion to U.S. AI infrastructure over three years, impacting decentralized AI narratives and crypto markets.

Meta is pivoting to a closed-source AI model codenamed 'Avocado,' expected in spring 2026. Led by TBD Lab, it integrates third-party models like Google Gemma and Alibaba Qwen. Meta reallocates resources from metaverse to AI, planning $60 billion investment in U.S. AI infrastructure over three years, potentially impacting crypto AI sectors.

Meta abandons open-source AI with new closed-source model 'Avocado' (spring 2026) led by TBD Lab, integrating Google Gemma and Alibaba Qwen. The company reallocates funds from metaverse to AI hardware with a $60 billion U.S. infrastructure plan, impacting AI tokens and metaverse coins.