Something strange happens when you extrapolate current AI progress 20 years forward. The economics stops making sense.

Not in the way economists usually mean, where models break down at the margins. The whole framework stops working. Supply, demand, scarcity — these concepts assume things that won’t be true anymore.

Back around 2004, outsourcing was killing jobs furiously, and people wondered if there would be any jobs left. There was a Business Week article "Where are the Jobs?". Well, the link is dead and it's not at the archive, so the PDF I saved came in handy (warning on link and article rot) [0].

My "answer" at the time was that FAB would allow everyone in the boonies (as I drove down I-5 in the middle of absolute **** nowhere) to create bespoke products that have personal, local, or group value.

I hear the call again for average Joe's to become artisans. I have no answer on the tip of my tongue, other than my service experience as an Electrician's Mate (with a physics degree, but that's another story) and until robots can wire homes, transmission lines, substations and data centers, right now, I tell all the young people looking for work to become electricians.

And I'm going to stop right here, because going any further needs some thinking through.

[0] https://pastebin.com/raw/Fi5YSV4a (text)

That's a nice scenario, but not the realistic scenario. The realistic scenario is that people will use designs from social media, and everyone will look rather similar.