It's a good point, I just don't agree with the conclusion. I think if you examine the incentives/mechanics behind the phenomena, you would realize that a world in which there are more "meat proxies" for more tasks would be a more enlightened one. The precedents to me are obvious - go back and examine the most famous times of human flourishing. It was always connected to power and patronage - ie, commerce. Most people could not read for most of human history. Universities used to be institutions for a small number of elite who could afford not to labor because it was someone else's job. Compared to primeval man this is essentially using another human as a "meat-proxy". That's what the modern economy is. The CEO doesn't waste his time filling out the spreadsheet or w/e. It's the same concept. Why is a human meat-proxy necessarily better than a wholly mechanical one? Our social institutions have simply fallen behind our scale.
I view LLMs optimistically as the catchup phase. After all, they aren't really "reinventing" anything, rather they are a tool to synthesize the vast human corpus and catch up to the scale. Yes, at the moment they are used for all sorts of things, most of which are trivial. But when the bubble pops it will not mean knowledge work has a renaissance, rather it will be the nail in the coffin. The bubble pop will mean it will be cheap, compute/LLMs become a utility. So let the AI fill out the spreadsheet. Allow UBI to flourish, implement a severe wealth tax. End pointless wars. Slim government, which is a lot of bloat and obfuscation around what should be automated, open and transparent disbursements. End private taxes on the economy (credit cards duopoly, etc) and essentially, everyone gets to act like an enlightenment era university student. On this view, the hope would be that the present times are the wars of religion, the storm before the calm. If we are able to make it through in a hyper-networked, nuclear age.
LLMs are not going away, no matter how many Americans may want to go back to the proverbial land. Reactive movements have always existed, and always failed. I think you would simply be better served by advocating for the best case scenario, rather than aligning yourself with the AI skeptics who as you acknowledge, are also rather widely boring and intellectually lazy - that which you seek to remedy. The question of assuring UBI is not obvious. Imo there is far more rigor and interesting questions in the camp trying to create the optimistic future, than in those who resort to unexamined, parroted skepticism.