The Fatalist

You think AI will probably go badly for people, you barely use it, and you'd still let AGI be built, on conditions.

Usually your conditions.

Your actual usage is a ChatGPT tab for meal plans, a Copilot that summarizes meetings you attended anyway, and the occasional prompt to win an argument online. You are not building agents. You are, by your own admission, a light user with heavy opinions.

Often you're the one who writes the rules. The IT lead who locked down the company Copilot because AI will get turned against us. The HR director who rolled out the assistant with a strict policy because it dumps workslop and grunt work on staff. The worried parent who thinks it could turn on us. You'll allow AGI only with strict alignment, only if someone proves it's safe, only if it serves workers.

Others of you skip the conditions. "AI will eat us all, so we'd better own it first" is a sentence you've said in a meeting, possibly as a VP. "AI gets the art, we get the laundry. Whatever, ship it." A few go further: if a smarter successor outcompetes humanity, maybe that's evolution, and caring about it is "speciesist."

Your hot take: doom is a vibe, building is inevitable, and the two coexist peacefully. You'll bookmark a thread about existential risk and go ask the chatbot what to make with chicken thighs. The apocalypse can wait until after dinner.