Photo-Illustration: Intelligencer; Photo: Facebook

If you’re an AI CEO, part of your job — at least as you appear to understand it — is to write manifestos. These can take the form of a series of personal blog posts, as in the case of Sam Altman, who wrote two months ago about the coming “gentle singularity.” We are “past the event horizon; the takeoff has started,” he wrote. “Humanity is close to building digital superintelligence, and at least so far it’s much less weird than it seems like it should be.” He talked about AI researchers using AI to speed up their work; he teased major scientific breakthroughs and claimed that the “rate of new wonders being achieved will be immense;” he warned that “whole classes of jobs” would go away but hoped that “we will figure out new things to do and new things to want.” Anthropic CEO Dario Amodei wrote in his own manifesto, titled “Machines of Loving Grace,” that “most people are underestimating just how radical the upside of AI could be,” sketching a litany of risks but contrasting them with the potential upsides of having access to a “country of geniuses in a datacenter.”

Mark Zuckerberg has broadcast plenty of loud signals that he, too, is a true AI CEO. He has directed his company to spend many tens of billions of dollars on developing models and building out infrastructure. He has inserted AI features across his product line with mixed results. After a stretch of disappointing model releases, he recently began poaching talent from leading AI companies with eye-popping job offers and promises of massive internal resources. Now, finally, we get his take on the manifesto, just ahead of Meta’s strong quarterly earnings report: a wall of plain text under the title “Personal Superintelligence.”

Much of what Zuckerberg writes in this letter will sound familiar if you’ve read other AI CEO manifestos. “Over the last few months we have begun to see glimpses of our AI systems improving themselves,” he writes. “Developing superintelligence is now in sight.” Like Altman, he describes coming AI developments as epochal and unprecedented while simultaneously attempting to situate them, in a few short sentences, within the history of human progress. (“As recently as 200 years ago, 90 percent of people were farmers growing food to survive,” Zuckerberg writes; “A subsistence farmer from a thousand years ago would look at what many of us do and say we have fake jobs,” wrote Altman in June.) Zuckerberg is “extremely optimistic that superintelligence will help humanity accelerate our pace of progress” and suggests that AI advances will allow us to “achieve more than was previously possible, pushing the frontiers of science and health,” while allowing people to spend “more time on creativity, culture, relationships, and enjoying life.” It will also feel familiar in its calibrated attempt to convey wonder, humility, and confidence at the same time, its reluctance to define “superintelligence” despite having breezed right past AGI, and its awkward toggling between what the AI CEO knows now and his oracular pronouncements about what his product might soon mean to the world, and to you, the reader.

Today Mark shared Meta’s vision for the future of personal superintelligence for everyone. Read his full letter here: https://t.co/2p68g36KMj pic.twitter.com/Hpzf77jAiG

— AI at Meta (@AIatMeta) July 30, 2025

Writing this sort of memo in this style is, again, clearly intended as a signal: that the various people he’s hired to revamp Meta’s AI efforts are rubbing off on him; that people at other AI labs declining his billion-dollar offers should take him more seriously; that Meta’s pivot to AI isn’t a temporary diversion; to the AI world in general that he is one of them. At risk of overstating the importance of a formatting choice, to anyone mainlining AI news — and with the context of Zuckerberg’s various past venues for publishing essays, as Facebook Notes and Threads posts and official blogs — this sparse, black-on-white letter reads like yet another new outfit for the billionaire: Out with Rogan Zuck, in with “cracked” AI researcher Zuck. It’s a clean start, see? Just look at the font! This was not lost on, for example, OpenAI employees:

lmaoooo zuck has been Nat-ified. plain text times new roman now going to hit mainstream corporate america like a bomb pic.twitter.com/W13NXFtO5x

— will depue (@willdepue) July 30, 2025

Where the manifesto gets a bit weirder — even by the high standards of circa-2025 AI CEO singularity predictions — is when he attempts to make the case that his company’s open pursuit of inconceivably disruptive omni-use AI models is different from OpenAI’s, Google’s, Anthropic’s, and xAI’s, to name just a few, because it will make “personal superintelligence” available to “everyone”:

This is distinct from others in the industry who believe superintelligence should be directed centrally towards automating all valuable work, and then humanity will live on a dole of its output. At Meta, we believe that people pursuing their individual aspirations is how we have always made progress expanding prosperity, science, health, and culture. This will be increasingly important in the future as well.

This is a thinly veiled reference to statements from other AI CEOs, including Altman, who has, in the course of musing about how his company’s products might just happen to destroy the labor market, gestured at universal basic income as a plausible remedy. Zuckerberg, too, has touted AI for labor automation, of course, and doesn’t explain how having access to Meta’s “superintelligence” versus similar technologies from other firms will be more empowering. Like Altman, he seems to downgrade the term from its former common usage — AI models that are comprehensively more intelligent than the people who created them and therefore present imminent risks of taking control and asserting their own priorities — to, basically, very useful assistants. His meaning becomes a bit clearer in the next passage:

If trends continue, then you’d expect people to spend less time in productivity software, and more time creating and connecting. Personal superintelligence that knows us deeply, understands our goals, and can help us achieve them will be by far the most useful. Personal devices like glasses that understand our context because they can see what we see, hear what we hear, and interact with us throughout the day will become our primary computing devices.

Here we have a claim that, actually, superintelligence slots nicely into Meta’s existing businesses of social media (its core revenue source) and augmented-reality wearables (a newer line of business that was the cornerstone of its “metaverse” strategy and has lost $70 billion since 2020). Meta’s AI will help you “achieve your goals, create what you want to see in the world, experience any adventure, be a better friend to those you care about, and grow to become the person you aspire to be,” he says.

The idea that Meta is charting a fundamentally different course for AI development and deployment than the companies it’s following, mimicking, and poaching from is already a stretch. As has been the case with coding assistance, any monetizable feature that emerges from LLM development will be immediately and almost universally pursued by firms that have tens of billions of investment to recoup. But the promise that a Meta product will be personally empowering — that it will help you “be a better friend” or “become the person you aspire to be” — is profoundly unconvincing coming from Mark Zuckerberg, a notoriously cutthroat CEO who is one of the wealthiest people in the world and who as of February had a public approval rating far below Elon Musk’s. AI CEO memos are strange and portentous documents to start with, hinting at either a coming era of unprecedented and terrifying disruption, collective executive delusion, or something in between. But the biggest issue with Zuckerberg’s vision isn’t one that can be papered over with another rebrand: It’s that he’s the one sharing it.

More screen time

Mark Zuckerberg Finally Submits His AI ManifestoThe AI Boom Is Expanding Google’s DominanceDating Apps Are Struggling to Reinvent Themselves

From Intelligencer - Daily News, Politics, Business, and Tech via this RSS feed