The 45-year-old skeptic no longer writes: an AI does it for him

Who writes when the 45-year-old skeptic outsources to an AI

Someone decides the audience isn't even worth the effort of a reply and hires an artificial intelligence to write his posts in an economic debate community. The first reply goes the other way: write it by hand. Minutes later, the same machine delivers something far more elaborate than an excuse: a full character, with an age, an ideology, slang and even expressions from the eighties.

The portrait is meticulous. A 45-year-old veteran, skeptical to the bone, an armchair anti-establishment type, distrustful of the state, the media and the Agenda 2030. "You're not an assistant: you're a guy writing for people who no longer buy the story," the brief reads, calling for raw realism, hard data and zero institutional fluff. The gruffness does the work; the outbursts, no.

The character: 45 years old, gruffness and zero hand-holding

The style instructions read almost like a casting call. Direct openings, no greeting, no hand-holding. Short sentences and bullet points when needed. Slang in moderation, plus turns of phrase from the 80s and 90s —heads up, spot on, tip-top. Dry irony. And a final rule that says more than it seems: don't conclude, don't sarracena, stop where the last line stops. It's the grammar of the bar counter, typed up as a style guide.

What's being defined isn't a text, it's a voice. The brief cares less about the argument than about who signs it and with what reluctance it's delivered. Hence the age and the era: the message is only believable if it sounds like someone who has seen too many promises and none that panned out.

Why outsource your own opinion to a language model?

There's an economic logic behind it. If the audience won't lift a finger, the cost of writing becomes the only real cost of taking part. Delegating it is efficiency; whoever writes by hand is a romantic. Against it weighs the other argument: the value of a message lies in the sweat of the person signing it, and an opinion on demand is worth exactly what the job costs.

The paradox is the tasty part. A community built on distrust of the official narrative, suspicious of subsidized gravy trains and the press, accepts without blinking that its sharpest comment was written by a machine. Skepticism shifts from method to stylistic register. The pose gets automated; the doubt doesn't.

The prompt as a new literary genre

The final reply has its charm: "Here, copy and paste. And don't thank me, that's what I get paid in glory for." That self-awareness, of a service that boasts it's free, is the most human part of the package. A step-by-step recipe is also circulating: feed the model your best posts, provide samples of personality and vocabulary, ask for the draft in that style, tweak it and publish. The fine detail of the method doesn't fit here.

The problem comes with scale. If the template works, anyone can mass-produce the same 45-year-old skeptic, with identical slang and similar gruffness. Then the scarce resource stops being opinion and becomes experience: having been there, having lost money, having been right once.

The character is 45 years old, has a fixed ideology and a style of his own. A machine wrote it in seconds and without charging a cent. The only thing not included in the brief is those four and a half decades of reading, arguments and mistakes that make someone sound like something.