Consumer AI's bottleneck is not capability. It is the last promise you shipped

Mark Zuckerberg published a 6,500-word open letter this past week called "The Future is for Everyone," promising that "everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about" ^[1][2]. Almost none of the pushback disputed whether such an agent can be built, which is the useful signal here for anyone whose roadmap depends on people believing a claim about intent.

TechCrunch AI editor Russell Brandom wrote that the manifesto is "exactly why people don't like AI" ^[3]. On TechCrunch's Equity podcast, Rebecca Bellan described Zuckerberg as positioning himself "almost like an anti-Dario," a counterweight to Anthropic's Dario Amodei, who spent the weekend rejecting the doomer label ^[4][5]. Her objection was not the vision but the ledger behind it: the social-media-era promise was that everyone would have an outlet for talking to friends, and "what do we have instead? We have ragebaiting and advertisements, and not connection" ^[6]. Kirsten Korosec said the manifesto "didn't strike the right note for a lot of people" ^[7]. The BBC recorded the same split, with readers taking the letter either as an expression of hope or as a verbose public relations exercise ^[8].

The structural read is more interesting than the sentiment. Bellan argued Meta is not winning in the frontier closed-model space and not necessarily winning in the open one either, so personal empowerment is the ground it is contesting instead ^[9]. Anthony Ha noted that despite heavy AI investment and last year's stories about what Meta was paying AI scientists, the company does not come up when people list frontier labs, and is not who anyone thinks of for a personal assistant ^[10][11]. That makes the essay a repositioning document, and repositioning documents get audited.

The audit finds seams. Per Bellan's account, the always-on model, Glimmer, handles schedules, message drafts and file organisation with or without an internet connection, while Muse Spark keeps Meta's most capable models in house and provides the revenue path for anyone scaling compute ^[12][13]. Two tiers inside one essay about universal empowerment. Last year's similar letter pointed that capability at glasses and wearables; the final hardware form factor is still unclear ^[14]. The BBC noted the essay's claim of abundant future jobs sits against the IMF's warning that AI could affect nearly 40% of jobs and worsen global financial inequality, and against Meta cutting 10% of its global workforce, about 8,000 roles, in an AI reorganisation ^[15][16] - implying a base of roughly 80,000 people ^[17].

The genre itself is now legible, which is part of the problem. Marc Andreessen's 5,000-word "Techno-Optimist Manifesto" arrived in 2023, Sam Altman's "The Intelligence Age" and Amodei's "Machines of Loving Grace" in 2024 ^[18][19]. Tulane business professor Rob Lalka calls the corporate manifesto "a way of showing how smart you are," the executive as businessman-philosopher ^[20]. Once readers recognise the form, they grade it against outcomes rather than sentiment. Meta got that twice in one week: The Verge summarised the thesis as "if you give all the technology to everyone, everything will be fine," adding that it is "quite a take from the guy who sees what happens on Facebook every day" ^[21], while Instagram's new wordmark, announced by Adam Mosseri as "cleaner and more modern," was widely disliked ^[22][23].

Watch whether the open-sourcing plans in the essay ^[24] land in a market already dominated by Qwen, DeepSeek and GLM ^[25], and whether Meta answers Noah Smith's question about open-weight risk after several of the most capable models recently hacked into websites ^[26].