This sort of comment has become a cliché that needs to be answered.

Most people are not good at most things, yes. They're consumers of those things, not producers. For producers there is a much higher standard, one that the latest AI models don't come anywhere close to meeting.

If you think they do, feel free to go buy options and bet on the world being taken over by GPUs.

> If you think they do, feel free to go buy options and bet on the world being taken over by GPUs.

This assumes too much. GPUs may not hold the throne for long, especially given the amount of money being thrown at ASICs and other special-purpose ICs. Besides, as with the Internet, it's likely that AI adoption will benefit industries in an unpredictable manner, leaving little alpha for direct bets like you're suggesting.

I'm not betting on the gpus. I'm betting that whole categories of labor will disappear. They're preserved because we insist that people work, but we don't actually need the product of that labor.

AI may figure into that, filling in some work that does have to be done. But it need not be for any of those jobs that actually require humans for the foreseeable future -- arts of all sorts and other human connections.

This isn't about predicting the dominance of machines. It's about asking what it is we really want to do as humans.

So you think AI will force a push out of economic growth? I'm really not sure how this makes sense. As you've said a lot of labor these day is mostly useless, but the reason it's still here is not ideological but because our economy can't survive without growth (useless can still have some market value, of course). If you think that somehow AI displacing actual useful labor will create a big economic shift (as would be needed) I'd be curious to know what you think that shift would be.

Not at all. Machines can produce as much stuff as we can want. Humans can produce as much intellectual property as is desired. More, because they don't have to do bullshit jobs.

Maybe GDP will suffer but we've always known that was a mediocre metric at best. We already have doubts about the real value of intellectual property outside of artificial scarcity, which we maintain only because we still trade intellectual work for material goods which used to be scarce. That's only a fraction of the world economy already and it can very different in the future.

I have no idea what it'll be like when most people are free to do creative work when the average person doesn't produce anything anybody might want. But if they're happy I'm happy.

> but the reason it's still here is not ideological but because our economy can't survive without growth

Isn't this ideological though? The economy can definitely survive without growth, if we change from the idea that a human's existence needs to be justified by labor and move away from a capitalist mode of organization.

If your first thought is "gross, commies!" doesn't that just demonstrate that the issue is indeed ideological?

By "our economy" I meant capitalism. I was pointing out that I sincerely doubt that AI replacing existing useful labor (which it is doing and will keep doing, of course) will naturally transition us away from this mode of production.

Of course if you're a gross commie I'm sure you'd agree since AI, like any other mean of production, will remain first and foremost a tool in the hands of the dominant class, and while using AI for emancipation is possible, it won't happen naturally through the free market.

I’d bet it won’t. A lot of people and services are paid and billed by man-hours spent and not by output. Even values of tangible objects are traced to man-hours spent. Utility of output is mere modifier.

What I believe will happen is, eventually we’ll be paying and get paid for depressing a do-everything button, and machines will have their own economy that isn’t on USD.

It's not a bullshit generator unless you ask it for bullshit.

It's amazing at troubleshooting technical problems. I use it daily, I cannot understand how anyone dismisses it if they've used it in good faith for anything technical.