Mark Zuckerberg creating Meta Superintelligence Labs. Read the memo

Mark Zuckerberg said on Monday. Meta Superstelligigence laboratories, some of the latest recruitment, including Skala AI former Alexandr Wang and former Github CEO Nat Friedman.

Zuckerberg said that according to an internal note obtained by CNBC, the new AI Superintelligige unit will work on basic models such as open -source software, products and basic artificial intelligence research projects.

Bloomberg first reported About the new unit.

Meta’s founding partner and CEO, OpenAi and Google. At the beginning of June, the company said that Meta will now hire Wang as a part of an investment of $ 14.3 billion to the Scalated AI, the META’s Chief AI officer and some colleagues.

Meta also hired Friedman, CEO of Safe Superintelligigence, CEI, CNBC’s CNBC’s Founding Partner Ilya Sutskever, and Friedman, CEO of Safe Superintelligigence, and business partner Daniel Gross. Meta had tried to buy a safe super -siper, but was rejected by Sutskever.

Openai CEO Sam Altman said that Meta was hired in a recent podcast AI Researchers It offers a high signature bonuses of $ 100 million from its company.

META Technology Chief Andrew Bosworth said that Openai opposed Meta’s proposals in an interview on June 20, CNBC’s “closing Bell more messages”.

“Sunday, as a technology manager, sets a ratio for an incredible and unprecedented ability level in my 20 -year career as a technology manager.” He said.

Here is Zuckerberg’s full internal note published on Monday:

As the speed of artificial intelligence progress accelerates, the development of super -siper emerges. I believe this will be the beginning of a new era for humanity, and I am completely determined to do what I need to pioneer Meta. Today, I would like to share some details about how we organize our AI efforts for our vision: personal super for everyone.

We will look for our general organization Meta Superstelligigence Labs (MSL). This includes a new laboratory that focuses on improving our new generation models as well as all our foundations, product and fair teams.

Alexandr Wang joined Meta as our Chief Officer and MSL leadership. Alex and I have worked together for a few years and I see it as the most impressive founder of his generation. He has a clear understanding of the historical importance of Superintelligigence and built Scalai as its founding partner and CEO to a fast -growing company in the development of almost all leading models throughout the sector.

Nat Friedman also joined Meta to become a partner with Alex to lead the MSL, chaired our work on AI products and practical research. NAT will work with Connor to describe his future role. He run Github in Microsoft and operated one of the last leading AI investment companies. NAT served in our commodity advisory group last year, so it has a good understanding of our road map and what we should do.

We also have a few powerful new team members who have participated in today or participate in the last few weeks.

• TRAPIT BANSAL He pioneered RL, the common creator and common creator of O series models in Ophaneai.
• Shuchao bi -GPT-4O sound mode and the common creator of O4-Mini. Openai’de Previously Multi -moded LED LED.
• Huiwen chag -GPT-4O’s creator of image production and Google Research’te previously invented Mascgit and Muse Metin Architectures
• Ji lin -O3/O4-Mini, GPT-4O, GPT-4.1, GPT-4.5, 4O-Imagegen and operator helped to create a pile of reasoning.
• Joel Pabar – I’m in anthropic. Previously in HHVM, Hack, Flow, Redex, Performance Tools and Machine learning for 11 years in Meta.
• Jack Rae -Reasoning for Gemini Bullet and Gemini 2.5 for Gemini. In Deepmind, Gopher and Chinhilla directed early LLM efforts.
• Hongyu Ren -GPT-4O, 4o-mini, O1-Mini, O3-Mini, O3 and O4-Mini. He previously led a group after training in Openai.
• Johan Schalkwyk – The former Google member, who contributed early to Sesame and technical leader for Maya.
• Pei Sun -Google Deepmind for twins after training, coding and reasoning. Previously, Waymo’s perception models had created the last two generations.
• Jiahui yu -O3, O4-Mini, GPT-4.1 and GPT-4o’s common creator. He previously led the perception team in Openai and was a joint multimodal leader in Gemini.
• Shengjia Zhao -Chatgpt, GPT-4, All Mini Models, 4.1 and O3 common creator. Openai has previously directed synthetic data.

I am excited about the progress we planned for Lama 4.1 and 4.2. These models are an increasing number of representatives that help improve our products and technology in Meta AI and Meta, which are used by more than 1 billion monthly active in our applications. We are determined to continue to create these models.

Parallel, we will start researching our new generation models to reach the border next year. I have promised the last few months to bring together the best people in meta, other AI laboratories and to bring together the founding group for this little talent intense effort. We are still building this group and we will ask a few people in AI to join this laboratory.

The commodity is uniquely positioned to offer superity to the world. We have a strong job that supports creating significantly more calculation than smaller laboratories. We have a deeper experience of development and growing products that reach billions of people. We lead and lead the very fast growing AI glasses and wearable categories. And our company structure allows us to act with much more faith and courage. I am optimistic that this new talent and parallel approach to model development will prepare us to fulfill our personal superity promise for everyone.

In the coming weeks, there are even more wonderful people who participate in this effort, so keep watching us. I’m excited to dive and start to work.

WRISTWATCH: Meta’s AI Talent Spending Madness.