[petercooper]

Plenty of others have predicted the rise of unemployment in proportion to increased automation, and they've been wrong before.

Have they? In the long term, definitely, but in the short term, a sharp increase in automation is sure to have an effect on the amount of human labor required. Isn't a normal cycle that technology begins to replace workers but then those workers retrain and then get new types of jobs and recover their employability? It seems we're currently doing a poor job training existing workers and future workers, if anything.

[ikarous]

I was referring to the long term, but you make a good point. Asimov posited that increased automation would temporarily increase unemployment but that the net benefit would always be positive. He thought that automation would allow human beings to focus on more interesting and rewarding jobs.

The unspoken assumption is that these more interesting jobs are plentiful enough to support the displaced workers assuming that they're willing to train and transition into a new kind of job. Is this the case today? Are the "higher" jobs plentiful enough? The question is not rhetorical; I'm asking because I can't say. I work in software development and my perspective is limited.

[petercooper]

Ok, cool, seems like we're thinking along similar lines :-) The problem I see, however, is perhaps different to the one you pose. As distasteful as this sounds, I suggest that perhaps the "average" person is ill equipped to deal with those more interesting and rewarding jobs, given that they're likely to require significant technical or creative skills.

Can we bring "the masses", as it were, up to the levels of talent and expertise we'll require in the future, or is there a limit to how good our education and training can be? Will we reach a point where a small number of people are, effectively, supporting the majority? A mini version of this seems to be happening in some Western countries with people who've lived their entire lives on welfare. Perhaps the future is just one where more of us end up on welfare and the government keeps us going on the spoils of the extremely smart and/or powerful.

[orangecat]

Perhaps the future is just one where more of us end up on welfare and the government keeps us going on the spoils of the extremely smart and/or powerful.

I think that's essentially correct. The non-singularity non-fiscal-collapse endgame is that everything we need and almost everything we want can be produced by a small fraction of the population, and most people don't have sufficient skills and knowledge to be economically productive. When you're trying to cure cancer or build self-driving cars, 10 brilliant PhDs beat 1000 average high school graduates.

Prediction: the idea of a universal guaranteed income (e.g. Milton Friedman's negative income tax) will steadily become more popular.

[rpwilcox]

Agreed - the idea of retraining "wrench turning" jobs to be computer programmers is, in my mind, optimistic. Because every startup wants at least a BS in some tech related field (so a 4 year barrier to entry)... in addition to the fact that it takes a lot more than 4 years of school to train someone to be a computer programmer (I think)

[VladRussian]

>He thought that automation would allow human beings to focus on more interesting and rewarding jobs.

i think he is right. There is a bunch of problems in theoretical physics that somebody need to attend to asap and specifically there are some in gravitational waves theory that i'd like to get into as soon as somebody automate the enterprise software development. In the meantime we can't find people to fill open positions (30%) in our department.