AI Is a UFO

Meta laid off 8,000 people because they saw a UFO.

That's not the official story. The official story is that the nature of work is changing teams are being "right-sized" for the new era. Read between the lines: a thing showed up that doesn't match any existing category, it appears to have zapped a few departments, and the response is to clear the area and try to figure out what it is.

Framing for AI gets the shape wrong. It's not the printing press. It's not the internet. It's not a productivity tool, or a junior developer, or a new department... Every metaphor breaks. We keep reaching for things we already understand, and this thing isn't that.

It hovers! It does some of what a person does, some a tool, and a few things neither has ever done, and something else you can't describe. Then it changes overnight.

If you treat AI like software, you build a roadmap and watch it get obsolete the next day. The shape doesn't fit any of the boxes.

This is why corporate leadership looks dazed. Most leaders are trained to handle predictable shapes — markets, headcount, capital, time, cost. The job is to take an opportunity, decompose it into workable chunks, and assign it to humans. AI does not decompose into work like that. The Venn diagram of things it can do, things it can't do, and things it can't do reliably yet rearranges itself constantly. Strategy meetings that used to take quarters now take days, and the leaders who can't operate at that speed freeze. Freezing looks like layoffs, especially at large orgs with expansive systematized workflows.

The move isn't to pretend you know the shape. It's to admit you don't!

Send one good person out to make contact with the thing and report back. Like an expedition! "Alright — go learn about that thing but don't fucking die. Come back alive." It's totally fine to come back reporting you didn't see anything. Or you did and it scared the shit out of you, so you ran.

Resist the urge to staff a VP of AI who promises a roadmap, because there is no roadmap for a UFO. There's only a posture: stay curious, and don't bet the company on what the saucer will do next.

Don't deploy a tool. Make contact.