Analysis
I've read more of these manifestos than I'd like to admit, and I think Fortune's framing undersells what's actually going on. The question isn't "who's reading a 6,500-word essay from a tech CEO" -- almost nobody reads the whole thing, and that's fine, because the essay was never really written for a general reader. It's written for three specific audiences: employees who need a shared narrative to rally around during a brutal talent war, investors deciding whether to fund a $2 trillion IPO or a $500 billion compute financing platform, and policymakers who are actively drafting AI regulation and want to know where a company stands before they write a rule that constrains it.
Dario Amodei's essays and Mark Zuckerberg's public posts this year both do real work on at least one of those fronts. Amodei's writing has consistently previewed Anthropic's positioning on AI safety months before the company's product or funding decisions caught up to it -- which matters more this week than usual, with Anthropic reportedly circling a $2 trillion valuation and needing every institutional investor in that syndicate to believe leadership has a coherent, defensible view of where the technology goes next. Zuckerberg's posts about open-weight models and pricing have functioned similarly: he was framing Meta's Muse Spark pricing as a value play against "extreme" competitor margins well before this week's Grok 4.6 launch made pricing pressure the actual news story.
“Dario Amodei's essays and Mark Zuckerberg's public posts this year both do real work on at least one of those fronts.”
The manifesto, in other words, isn't really content. It's a hedge. It lets a CEO stake out a position early enough that when the company's actual strategy catches up -- a pricing move, a safety commitment, an open-weight release -- it reads as consistent with something they said months earlier, rather than as a reaction to competitive pressure. That's a genuinely useful tool for managing how a company's story gets told, and it costs almost nothing to produce relative to the value of shaping that narrative ahead of a fundraise or a regulatory fight.
Room for disagreement: it's entirely possible I'm reading intentionality into what's really just CEO vanity and access to a large in-house comms team. Plenty of these essays are genuinely self-indulgent, wander well past the point of strategic usefulness, and get written because a CEO has the platform to publish 6,500 words and few people around them willing to say it's too long. Readership data, to the extent it exists, would probably show most of these essays get skimmed by press and a small number of engaged employees, not read cover to cover by the LPs and regulators I'm arguing they're aimed at -- and a essay nobody reads in full can't do the narrative-shaping work I'm crediting it with.
Either way, the volume of these essays is itself a signal: when three or four major AI labs are all publishing long personal philosophy pieces in the same stretch of weeks, that's a sign the industry's leaders think the public narrative about AI is genuinely up for grabs right now, not settled -- and none of them want to be the one who let a rival define it first.
