I think I’ve already done this song and dance but my take on technological unemployment is: general AI will probably have a catastrophic effect on employment but specific AI likely won’t, and it seems really unlikely current AI technologies will generalize to general intelligence. It could be just around the corner, but it’s a lot more likely that if you gear up for it now you’ll look like those guys at the dawn of computing talking about how natural language processing was right around the corner. Admittedly when we do have real artificial general intelligence there’ll only be like 15 years to decide how we want to handle it, but that may well not even occur in our children’s lifetimes, so it hardly seems worth trying to time it.
It does seem like we should anticipate large-scale short-term unemployment from future technological innovation, though, often of people who are good at the thing that’s obsolete, bad at the thing that replaces it, and – in many cases – too old or tied-down to start over. If current events show anything it’s “probably while awaiting a return to homeostasis you should try to make those people feel like the entire system isn’t out to fuck them".
The question, of course, being how you do this cheaply.
What makes me sad is that there’s decent bipartisan support for skills retraining and job transition services and stuff like that, but in practice any attempt to do that seems to result in 90% of the money being funded to garbage profiteers who are good at bidding for contracts and pitching services to the government.
Wage subsidies. Let the people work cheaply enough and the economy will find jobs for them.
