So, jobs are being replaced by AI spending (not by AI), and AI spending is increasing because of AI, which means AI is replacing the jobs. Did I get wrong?
I suppose the distinction is worth making, since if it was actual AI replacing jobs, i.e. AI that is capable enough, today, to do someone's job, or more likely to increase someone's productivity so that headcount can be reduced, then that would seem a permanent shift and is only going to ratchet up.
OTOH, if it is only dreams of AI, manifested as AI spending, or CEOs laying people off thinking that AI will soon (even if not today) be capable of backfilling them, then this may well backfire, and will be "reversed" if demand is less than forecasted and/or job-replacing AGI doesn't materialize, and all we get is productivity tools, useful mostly for a narrow band of jobs.
Incidentally, I think Karpathy might be right that there is pretty much only one job that seems it actually could be replaced by AI today, at least potentially, which is call center tech support, or customer support, staff, who are dealing with a narrow domain and just reading off a script.
Spending is happening only because, it promises to replace workers. And that possibility arises only because AI exists. Spending is only a means, not a cause.
Sure, but is the cause "AI" (here today, doing someone's job), or "promise of AI" (that may never materialize anytime soon, either in quantity or capability).
