Mark Zuckerberg releases 6,500-word manifesto, and the network mostly rolls its eyes

Meta’s chief executive published a lengthy document titled “The Future Belongs to Everyone,” outlining an optimistic AI-driven future in which each person has a distinctive personal assistant that understands them, their goals and everything that matters to them. In the latest episode of TechCrunch’s Equity podcast, Kirsten Korosec, Rebecca Balan and Anthony Ha dissected why the vision did not exactly capture the audience, including colleague Russell Brandom, who wrote that the manifesto is “exactly the reason people don’t like AI.”

Balan likened Zuckerberg to Anthropic CEO Dario Amodei, who spent the weekend defending AI against the “dumb-AI” label, and said Zuckerberg positions himself “almost like anti-Dario.” She argued that the problem lies in Zuckerberg’s and Meta’s history: “In the early days of social networks he said he wanted to make sure everyone had a platform to talk to friends. What did we get instead? Ragebait and ads, not connection.”

According to Balan, Meta is trying to win in a different arena: not on the closed-model front nor the open-source front, but on “personal empowerment.” The new Glimmer model is aimed at calendar management, message drafting, file organization and continuous operation with or without an internet connection. At the same time, Muse Spark keeps the company’s control over the most powerful models and creates a revenue stream from users who want to scale compute for larger projects. A year ago Zuckerberg promised in a similar letter that AI would reach glasses and wearables; this time the final form remains open.

The article also notes Meta’s sizable AI investments, including aggressive hiring of researchers, yet the company does not appear in conversations as a “frontier lab.” On the consumer side, people use Meta AI tools on Instagram and Facebook, but the firm is not seen as the go-to for a personal assistant. In that sense, the letter is an attempt to rebrand and claim territory where Meta has previously fallen short.

Balan conceded that Llama is an excellent open-source tool, but at the consumer level Meta AI usage boils down to “pretty crappy chatbots.” Brandom’s bottom line: the letter and the personal-empowerment idea leave a bad taste in the mouth, because the one who says them is Zuckerberg himself.