
Bryan Johnson, writing on X:
AI progress will fracture society. It will be too much change, too fast, for society to metabolize in an orderly fashion. Our identities, societal institutions, and shared narratives will fracture. A civilizational dissociation.
He then spends 800 words arguing that when everything collapses, the only rational objective left standing will be survival — which, conveniently, is the brand he's been building for five years. "Don't Die" isn't just a philosophy, it's a company, a content strategy, and a supplement line.
The trick here is that the opening diagnosis is sharp. Institutional adaptation is losing the race against AI-driven change. Narrative coherence is fracturing. The observation that people can't function without a shared story — and that AI is dissolving the old ones faster than new ones form — is worth taking seriously. Johnson reads the room well.
But watch the sleight of hand. He moves from "society will struggle to adapt" to "survival becomes the only rational objective" without arguing the case. That's not a conclusion. It's a launch sequence. Societies have metabolised the printing press, industrialisation, and nuclear weapons — messily, painfully, but through politics and institution-building, not by retreating to a single primal drive.
The entropy framing deserves special attention. It sounds like physics. It is not physics. Entropy is a precise thermodynamic concept. Using it as a metaphor for death, then treating the metaphor as a measurable system ("Don't Die is capitalism, with continued existence as the only acceptable profit") is rhetorical laundering — borrowing the authority of science to dress up a brand proposition.
The summary paragraph is the tell. Johnson writes: "No one knows what's coming with AI." And yet the entire preceding argument is a detailed prediction about what's coming with AI — civilisational fracture, the dissolution of power structures, survival as universal imperative. You cannot build a confident prescriptive framework on a foundation of admitted ignorance and then sell the framework.
Well — you can. He just did.
Stay current weekly
Get new commentary and weekly AI updates in the AI Primer Briefing.