Related: sources · notes · metadata · Published Pieces

The AI News Economy Will Dwarf the AI Persona Economy

Persona AI will be intimate. News AI will be infrastructural.

Private companion AI will be large. It may even be enormous. People will talk to models for advice, comfort, coaching, therapy-shaped reflection, tutoring, brainstorming, shopping, dating help, and everyday assistance. The human appetite for being listened to is not small.

But the AI persona economy is bounded.

It is private, inward-facing, and usually one user, one model, one thread, one relationship. Even when emotionally intense, it does not necessarily become public infrastructure. It may change people, comfort them, manipulate them, or become a massive consumer category. But it is not the largest surface of AI’s contact with society.

The larger surface is shared reality.

Every institution depends on public cognition. Markets depend on what investors believe. Elections depend on what voters believe. Wars depend on what citizens, soldiers, allies, enemies, and markets believe. Courts depend on evidence and narrative. Science depends on claims, replication, prestige, and funding. Public companies depend on guidance and interpretation. Universities depend on authority. Media companies depend on trust. Political movements depend on framing. Brands depend on identity. States depend on legitimacy.

AI enters all of that.

The AI news economy will dwarf the persona economy because it mediates the world people share. It is not just news in the old newspaper sense. It is the entire public layer of claims, counterclaims, explanations, frames, evidence, corrections, predictions, and reputations.

Every actor with an incentive to shape perception will use AI. If persuasive text, audio, video, analysis, microtargeting, narrative testing, and fake consensus become cheaper, they will be used. Campaigns, governments, companies, activists, scammers, intelligence agencies, media firms, founders, and influencers will use them.

The cost of propaganda falls. The need for verification rises.

That is the opening.

The next media layer is not faster summaries. Summaries are cheap. The next media layer is provenance-native public cognition: who said what, when, based on which sources, with what prior record, under what incentives, contradicted by whom, corrected when, and reused later by which artifacts.

The persona economy asks: can AI talk to me?

The news economy asks: can AI help me know what is going on?

That second question is larger because it is collective. A private assistant can satisfy one person’s need in one moment. A public cognition system can become upstream of journalists, investors, policymakers, analysts, educators, creators, and citizens. Once a provenance graph exists, it becomes a shared reference object. People can cite it. Agents can search it. Writers can build on it. Critics can challenge it. Track records can accumulate.

This is why the automatic newspaper matters. It is not a chatbot that answers news questions. It is a public memory system that makes claims accountable over time.

Automatic radio is its audio projection. People may not read a 20,000-word evolving brief, but they will listen to an intelligent traversal of it while walking or driving. They can interrupt, ask for sources, hear a real human voice clip, request the opposing frame, or publish their own response. The private experience remains easy, but the public object keeps compounding.

Persona AI will be intimate. News AI will be infrastructural.

The persona economy may own moods. The news economy will own reality.

That is why the latter is bigger.