[cvwright]

The main result that I’m thinking of is quite old — going back to the Ford assembly line days IIRC. NPR had a great segment on it several years ago. I’m on mobile now, but I’ll try to so some digging when I get back to a real computer.

I suppose there’s a valid question about how well those oldest results transfer from manual labor over to knowledge work like programming. The NPR guest that I’m remembering seemed to think that the effect might actually be even stronger for knowledge work, not weaker.