AI titan Meta appoints Shengjia Zhao, the co-creator of ChatGPT, as the Chief Scientist in the AI Superintelligence division
Meta, the parent company of Facebook, Instagram, and WhatsApp, is set to revolutionise the AI landscape with its newly established division, Meta Superintelligence Labs (MSL). The lab's primary focus is on developing advanced personal AI agents designed to assist people in thinking, creating, and decision-making in real time.
By 2026, MSL will have access to Meta's colossal 1 gigawatt cloud computing cluster, Prometheus, located in Ohio. This energy-efficient cluster is enough to power more than 750,000 homes, providing the necessary computational power for the lab's ambitious projects.
Mark Zuckerberg, Meta's CEO, has announced Shengjia Zhao as the Chief Scientist at MSL. Zhao, who co-founded MSL and has been its lead scientist since its inception, does not have a traditional research background. However, his contributions to AI, particularly in OpenAI's reasoning model o1, have been significant.
Shengjia Zhao is working under Alexandr Wang, the former CEO of Scale AI, who now heads the superintelligence unit at MSL. Meta has recently invested $14.3 billion in Scale AI, bringing Wang onboard.
MSL is also recruiting top AI talent, including Jiahui Yu, Shuchao Bi, Hongyu Ren, Trapit Bansal, and three employees from OpenAI's Zurich office. The lab aims to build a dense, top-tier team for advancing AI research and engineering.
The lab's key focus areas include developing the next generation of AI models with enhanced reasoning capabilities and real-time assistance, recruiting elite AI researchers and talent, investing heavily in cutting-edge AI infrastructure, and changing how users interact with digital content and advertising.
Meta's investment in AI is substantial, with plans to invest "hundreds of billions of dollars" into AI compute infrastructure. This investment is evident in the significant capital expenditure surge, with $17 billion spent in Q2 2025 targeted at expanding AI infrastructure and capabilities.
The lab's focus on building next-generation AI models with unparalleled compute access and an elite team strongly implies advanced research in AI reasoning, real-time inference, and decision intelligence. While detailed technical specifics on AI reasoning models from MSL are not publicly disclosed, it is likely that MSL will focus on AI reasoning models, as Meta does not currently offer a competitor to OpenAI's reasoning model o1.
In summary, Meta's Superintelligence Labs is concentrating on creating powerful, personal AI assistants that integrate sophisticated reasoning to augment human thinking and decision-making, backed by top AI talent recruitment and heavy investments in AI infrastructure. The ultimate goal is to build "personal superintelligence"—AI systems that function as continuous, trusted assistants integrated deeply into everyday life, rather than merely automating isolated tasks.
- As Meta Superintelligence Labs (MSL) continues to recruit top AI talent, it is likely that their research will focus on developing artificial-intelligence reasoning models, furthering the advancement of technology in this field.
- With access to Meta's colossal 1 gigawatt cloud computing cluster, Prometheus, by 2026, MSL will be able to harness the technology necessary to create personal AI agents capable of real-time thinking, creating, and decision-making, enhancing artificial-intelligence capabilities overall.
