Zuckerberg assembled a dream team of top AI talent: what his plan to achieve superintelligence looks like.

For years, Meta was synonymous with innovation in open language models . Under the scientific leadership of Yann LeCun, the company paved the way for AI development in the West. However, its latest model, LLaMA 4, failed to achieve the desired impact.

Meanwhile, China is taking the lead with proposals such as DeepSeek, Qwen, and Ernie , driven by a state-run strategy that combines academia, industry, and technical advances in science and mathematics.

In this new environment, Meta decided not to give up any more ground . Unlike Apple and Amazon, which opted for partnerships with leaders like OpenAI and Anthropic, Meta embarked on an aggressive mass recruitment offensive.

In recent hours, Zuckerberg outlined his vision for the future of AI in a letter published on the Meta website. In the letter, he referred to the desire to create a "personal superintelligence" within everyone's reach and predicted that smart glasses will play such a key role that they will become our "primary computing device."

"Let this power be in the hands of people, so they can direct it toward what they value in their own lives. This is different from others in the industry who believe that superintelligence should focus on automating all valuable work , so that humanity can then live off what it produces," he explained.

And he added: "At Meta, we believe that people pursuing their individual aspirations is what has led us to progress, expanding prosperity, science, health, and culture . This will also be increasingly important in the future."

Beyond promising to bring superintelligence to everyone, he maintained that the safe development of the technology will be essential. He also commented that new concerns will arise along the way that must be rigorously addressed to mitigate risks, and spoke of "being careful about what we choose to release as open source ."

A few weeks ago, in a move that shook the foundations of Silicon Valley, Zuckerberg officially announced the creation of Meta Superintelligence Labs (MSL), with which he intends to lead the development of general AI. But, more importantly, he announced the lineup of experts who will be responsible for achieving this goal.

In this new technological chessboard, every move counts, and Zuckerberg has begun poaching key players from his closest rivals to strengthen his presence in this competitive arena. In recent months, he has recruited key figures from his competitors, forming an elite team in areas where he has historically been weakest: vision, audio, and human-computer interfaces.

Notable figures include talent such as Huiwen Chang (imaging, ex-OpenAI), Ji Lin and Hongyu Ren (language models), Jack Rae and Pei Sun (pre- and post-training at Google), Joel Pobar (inference, ex-Anthropic), Johan Schalkwyk (voice, ex-Alphabet), and Daniel Gross (CEO of SuperIntelligence).

But the most high-profile move was the $13 billion acquisition of Scale AI , along with its founder, young wonder Alexandr Wang , who will lead the superintelligence field. He will be joined by Nat Friedman , the former CEO of GitHub and head of AI applications.

The strategy speaks for itself: Meta seeks to integrate intelligent agents into its social platforms, transforming the user experience with virtual assistants and generative personas.

AI will cease to be a tool and will become content. This vision is similar to the failed attempt at the Metaverse, now with foundational models that interact directly with users and offer services via APIs.

In parallel, Meta maintains its commitment to open source , albeit with a freemium approach: basic models to attract users and advanced versions to monetize.

The question is whether it will be able to build a cohesive culture around this constellation of talent. OpenAI and Anthropic have demonstrated that leadership in AI requires more than resources: it requires a shared mission.

Can Meta articulate that vision from platforms designed to maximize attention? Can Wang lead such a diverse and young team without losing his way? The answers are still unclear. But what is clear is that Meta means business. And what's at stake is not just leadership in AI, but how we engage with digital.