Mark Zuckerberg not only wants Meta to lead the field of artificial intelligence. He wants to build an AI that leads you.
In a memo of employees watched by Gizmodo on June 30, Meta CEO announced the creation of Meta Ultra Smart Labs, a major restructuring of the company’s AI, a single, ambitious goal: to provide personal ultra-wisdom to everyone.
Forgot the chatbot. Zuckerberg’s vision is even more grand. He dares to bet that within a few years, AI will not only answer your questions or write an email. It will manage your schedule, anticipate your needs, run your home, help you make decisions, and even guide your career. Call it a human service powered by Meta.
“As the advancement of artificial intelligence accelerates, the developing UX is coming soon,” Zuckerberg wrote. “This will be the beginning of a new era of humanity, and I am completely committed to what I have done for Meta to lead the way forward.”
Zuckerberg memos to employees, according to sources, put forward vision for “super smart” departments and new employees
He said the current “unique position” of the Yuan
Essentially, sales promotion is “we have money, calculations and scale, and most competitors don’t have it” pic.twitter.com/qqqyeynhcj
– Rat King 🐀 (@mikeisaac) June 30, 2025
The move is seen as a direct challenge to competitors. Alon Yamin, co-founder and CEO of Copyleaks, a stolen-prosecutor detection platform, commented: “The launch of Meta’s Super Smart Lab is not just an announcement: Meta won’t be second in AI.” He added: “Meta and Mark clearly see it as a creation or resting moment for AI leadership.”
Bloomberg reported on the memo earlier Monday. After that, a meta spokesperson confirmed his content about Gizmodo.
To achieve this, Zuckerberg is forming an AI Avengers. New employees represent the main talent coup:
- Alexandr Wang, founder of AI Scale AI and one of the most connected operators in the AI supply chain, has joined Meta’s new chief AI officer.
- Nat Friedman, former CEO and elite AI investor at Github, is about to join AI products and application research.
- Shengjia Zhao, one of the co-creators of GPT-4, recently left Openai to join Meta’s new lab.
These are people who have power to hire, reputations quickly build and think. Zuckerberg said more big names are on the way.
Meta already has a solid foundation with Llama, an open source family of large language models. The latest version is now powering Meta AI, which the company claims has over 1 billion users per month on Facebook, Instagram and WhatsApp.
But this is only the first stage.
Zuckerberg now wants to build a new generation of models that can be represented by what insiders call boundaries. In short, this means artificial intelligence that can reason, plan, adapt and act without human guidance. If successful, this will not only make Meta compete with OpenAI or Google DeepMind; it will make Meta the world’s first truly universal personal intelligence home. This kind of AI is more than just answering your questions. It continues your life.
The Yuan did not start from scratch. Through Facebook, Instagram, and WhatsApp, the company already knows you better than most friends. Over the past 15 years, it has quietly built a behavioral map of billions of people, tracking who you talk to, what you see, say, and what you buy. Now, the treasure trove of intimate data is fueled for the next stage: not only serving you, but knowing enough about you to make your life more enjoyable than you.
In Tech Speak, Zuckerberg’s vision is for “agent AI” or AI that can act on your behalf. Imagine having a personalized intelligence that can always live on your phone, glasses and all devices, unlimited capabilities.
- You don’t schedule a meeting, it’s true.
- You have not organized a trip, it has been booked.
- You don’t know what job to apply for, it’s already editing your resume and mocking the interview.
- You don’t ask what you wear, it sees your calendar and arranges your clothes.
Meta has a unique position in providing this mass-market personal AI brain. Unlike smaller labs, it has a multibillion-dollar global user base, with more behavioral data and a large amount of computing infrastructure than almost any company on the planet. Moreover, Zuckerberg’s controlling shares has no shareholder that can slow down.
But this also raises urgent problems. Who controls this intelligence? What will it prioritize? What happens when the machine knows your desires better than you and the machine works for the company?
Zuckerberg believes that Meta can be trusted to build and deliver super wisdom for the public. But the next chapter of AI has nothing to do with clever apps or productivity hacks. This will be a battle for whom the brain programming program.
