Kai-Fu Lee argues in a TED video that AI will be hugely beneficial for society by automating away mind-numbing jobs and driving people towards more meaningful roles – such as those that require a high degree of judgement in the face of complexity (e.g. economists, CEOs) and those which require empathy (e.g. teachers, etc.)
By and large I agree. In fact, economists such as David Autor have argued for a long time that automation as a whole will lead to “job polarization”; AI probably widens the scope of jobs that will be impacted, but the logic is essentially the same. (Parents should take note – your kids need to do more than memorize those times tables.) However, this will be cold comfort to the workers of today who will simply not be able to adjust to the new landscape for the remainder of their working lives. I expect to see anti-automation movements, similar to Luddites against machines during the Industrial Revolution. GPUs and remotely-located servers are a bit hard to destroy with a wrench, though, so this opposition will come virtually – through lobbying and regulation.
I recommend taking a bit of time to watch this video – it expresses its viewpoint quite succinctly.