While Ned Ludd's ghost might upvote this, I am as unconvinced of this as apparently everyone in this thread has always been of bitcoin.
For starters, any increase of output with a decrease of input labor is by definition an increase of productivity, individual, net, all of it.
For two, AI hasn't proven anything beyond quickly doing grunt labor. Quickbooks & Excel didn't put the CPA trade out of business, an algorithm that identifies a cat in a picture or whatever reduces overhead, not eliminating professions.
Tangenitally I am, however, very excited at AI's prospects in reverse-engineering and rewriting code, a long-lived fantasy that we can actually un-fark many old or particular programs and systems that are currently hardware dependent or just not worth the man-hours to do the transcription from, say, 16 bit to 64 (or 128, or 256, wherever the future goes).
Lastly, robots have been around since the 60s, industrially and standardized. They do good work, consistently, reduce cost and thus prices and floors of production. The automotive industry is bigger than ever, and more top heavy. This is good, no?
No-one dreams of working on an assembly line for Ford, and unlike the common-wisdom ideal of teenagers & inexperienced workers starting in the service industry and working up & out to better professions, I don't know of anyone who realistically has a career plan of 20 to 60 working their way up from floor sweeper to head of district in any enterprise.
There is this weird pessimism that comes on the tails of this breakthrough in labor leverage, that we're going to put ourselves out of work and have to sit around and then starve to death or get UBI because we're incompetent, when 100 years ago the plurality of Americans were subsistence farmers. Society alters to meet social needs, what we're really talking about is introducing a ton of slack into industrial labor. That isn't a waste, it's the potential for great expansion in so many other fields (think what resources were sunk into industry during Cold War that then turned to softer industries after 1989. Maybe the fantasy here is we should have the best and brightest of us, I dunno stay as mechanical engineers and work for GE or UTC building better stealth bombers instead of looking to FAANG and writing better ad service algorithms).
More energy is always a better thing for society. Labor saving tech is equivalent to increasing the sum of energy available to society. I don't cry because the ladder of the old gold union benes at John Deere were pulled up while I was in gradeschool, I went in search of a better opportunity than the one that wasn't there. The difference may be that "this will affect millions simultaneously!" but so hasn't every transformative technology.
Last-lastly, to make a general economic argument I never see: tens of millions without labor drives up supply which drives down wages. Robots and AI make everything for a fraction of cost (comparable to shipping manufacturing to slave, I mean indigenous, labor overseas for pennies-on-the-dollar in 1970s), so reduced wages maintain balance with cost of goods & services. Entrepreneurs can now exploit a larger labor force with diverse skills to explore new ventures with reduced costs on all front due to labor-saving tech and labor-bloated workforce. This leads to new ideas and extrapolations previously unviable or with a too-high floor of cost to entertain. Trillionaire tech CEOs who own everything divvy out grants like feudal lords funding weird idea-people like Leeuwenhoek with his curious "micro-scope".
Not ideal but that's still more than a starved populace voting demagogues for bread doles.