Future History of AI is a blog about reading science fiction as if it were research. Not the fiction we write next — the fiction we already have. For a century, the best writers in the genre thought hard about artificial intelligence: what it would want, how it would fail, who would be harmed, and what we'd trade away to have it. A lot of that thinking turned out to be uncannily accurate. The rest is a warning we can still act on.

Every post here does two things:

• Applies the lesson to now. A famous story isn't just a plot — it's an argument about technology and people. I unpack that argument and hold it against what's actually happening in AI today: the systems being shipped, the incentives behind them, the failures already showing up.
• Reads it forward. If the story got something right about the shape of the future, I follow that thread — what it implies about where AI is heading, and what we'd be wise to watch for.

So expect close readings of the canon — Asimov and Clarke, Dick and Gibson, Blade Runner and Her and Ex Machina — used as lenses on the present, not as nostalgia. The point isn't that the writers were prophets. It's that they were careful, and careful thought about the future is exactly what we're short on while it arrives.

Thanks for reading. New analyses posted regularly.