but people won't be out of a job, just out of old jobs done in old ways, to be replaced by new, more productive jobs done in new ways
How much analogous will that be? Beyond (maybe) certain 'general' largely-manual jobs that would be too expensive to replace with robots initially, it seems to me that there's no activity that humans would have in which they'll produce a higher value than even HLAI. And human-caring, initially at least.
At some point AI becomes powerful enough that it's no longer economical to employ humans. That's important, but it's also something like the next phase after the phase we're entering with AI.
The phase we're entering now is one where AI will automate and transform work in ways that make humans more productive. It's a bit unclear how long this period will last. My guess is between 15 and 30 years, because it'll take about that long for us to grow the economy enough to be able to afford to build AI powerful enough to fully replace humans. This is an often overlooked concern: we don't build AI just because we can, but because it's economical to. We've seen similar patterns during industrialization: we don't build a factory or automate something until it becomes cheaper than just paying a human to do it in a bespoke way.
So eventually, yes, AI eats everything, but there's likely enough years before that when we'll have to live through a world thoroughly transformed by AIs working with humans as productivity multipliers.
I have two motivations in this post:
Thus I'm going to highlight what I think are important analogies between industrialization and computerization and what I think that can tell us to expect from transformative AI in some ways. I want to be clear this is not very careful work so there's lots of caveats and exceptions and such. My aim is to 80/20 the analogy and the explanation.
I've not said anything here about risks, though note that the assembly line enabled the mass production of weapons that were way more deadly than what had previously been available. We should expect similar from transformative AI.
(Final note: I think there's some irony that I didn't use GPT-4 to help me write this as a real essay with references and nice paragraphs. Let's chalk this up to me still learning to use LLMs effectively in my work and using one sometimes imposing enough mental overhead to figure out how to do what I want that I prefer to do the lazy thing and not use it, though I expect future iterations of the technology to enable me to be this lazy and easily produce good results.)