AI and leisure time
Yesterday I offered, as one of nine reasons why AI wouldn't be leading to noticably more and better software for everyone, that efficiency gains might instead be captured by people making the software:
- Software teams are capturing a lot of the speedup in the form of more leisure or quasi-leisure time. (To greatly simplify: software teams are offering 2x speedups to their companies. They're then using tools that allow them to work 10x as efficiently, and taking nice long coffee breaks.)
It turns out a lot of people really dislike this idea. Understandably so! Here are some further notes:
- As I tried to make clear, this is one explanation among many. I do think this happens more than never (see below), but I definitely do not think that a lot of the AI productivity gains are going to programmer leisure time.
- But! What exactly is supposed to happen if a programmer or team (i) has access to AI tools, (ii) is structurally blocked from increasing output by more than (say) 2x, and (iii) doesn't want to quit? The extra time has to go somewhere. (Some of this might be not-very-glamorous forms of quasi-leisure that involve looking serious at your desk, worrying that you're going to get laid off, and so on.)
- The situation described above requires a parlay, is probably not very common, and is probably getting rarer, but the evidence I have strongly suggests that it does exist.
- I'd guess that being on Hacker News is negatively correlated with being in that type of workplace.
- As is probably obvious from the rest of this blog, I am extremely not seeing AI productivity increases turn into extra leisure time (except in the sense that programming is so enjoyable that it often feels like leisure). I have never spent so much time programming in my life, including (e.g.) when I was childless and working at an intense startup.