“Historically automation has replaced the jobs requiring the least skill with fewer jobs that required more skill”
This is a compelling narrative, but it’s not true. Technological unemployment has not historically accompanied increased levels of automation. There are some economists who project that this will change, but that would be a break with the past.
Probably worth remembering that the current wave of automation struggles to replace jobs we consider “low skilled”, because they actually require high skill (they are impossible to automate with current tech) and are performed for little money (so the capital investment in automation is uneconomic).
It most certainly is true! How many fields are tilled by hand today, instead of by tractor? How many jobs can be done today by someone who is illiterate? History is full of obsolete occupations.
I never claimed that mass technological unemployment had occurred historically, only that the minimum skill level required for employment has been steadily rising over the past ~150 years at an accelerating rate. The observation that "new jobs were always created in the past" fails to take into account the fact that the new jobs have _always_ required more skill than the old jobs they replaced.
Almost by definition, automation will perpetually struggle to replace the jobs we _currently_ consider low skilled. If they were easy to replace with current technology and expertise we presumably would have done so already. Those that remain are complicated enough not to be worth automating (yet), and thus remain the bottom of the barrel (for now).
That’s more than the entire global population was in 1964. So it’s not exactly like there’s a dwindling base of extremely high skilled jobs and majority unemployment. You are right that the minimum skill level is rising for many desirable jobs in developed countries. There are still plenty of jobs to be done that can be taught in one afternoon though.
It’s true that in many sectors in developed countries–such as agriculture–fewer workers use automation to achieve a higher level of productivity than historically. However, most industrial jobs lost in developed nations are a result of outsourcing to low cost economies, not being out-competed by automatons. The people who no longer have to toil in the fields didn’t just give up on life; they ended up as hairdressers, therapists, accountants, decorators, shopkeepers, software developers etc.
We’re not remotely close to automating vast numbers of jobs, and even when we do there will be more jobs to be done. Comparative advantage tells us that even if the machines are way better than people at everything, it will still make economic sense to put people to work at what they can do best.
This is a compelling narrative, but it’s not true. Technological unemployment has not historically accompanied increased levels of automation. There are some economists who project that this will change, but that would be a break with the past.
Probably worth remembering that the current wave of automation struggles to replace jobs we consider “low skilled”, because they actually require high skill (they are impossible to automate with current tech) and are performed for little money (so the capital investment in automation is uneconomic).