If you consider things like Mythos (I know it was partly hype), and cyber security using AI to find old vulnerabilities in open source tools, and others using that information to actively disrupt the economy (this is already happening)....
And governments pushing quantum computing, presumably to be first to crack Internet security: it's easier to imagine some of those future threats.
It's a massive long shot pipe dream, but we might somehow end up with legislators who have backbones and actually represent the people, and they'll enact regulation to reign in the applications of AI, provide a safety net for the millions who will be affected by it, and/or at least find a way to spread the prosperity that comes from AI across the population rather than into the hands of the few. I'm not betting on it, but it theoretically could happen.
I doubt it. There is so little care for monopoly or copyright because AI is a race, as much as it is a tool. And nobody wants to restrict the race. Policy may pass that limits the use of AI for the general public, but the 'whatever it takes' race to 'superintelligence' goes on.
> Or that a competitive startup won't lean on AI to get ahead?
Specifically, with the way both economy and politics is structured, everything will be about big corporations with centralized power. A competitive startup leaning on AI getting ahead will be either destroyed or bought.
>The world doesn't hold hands with anyone, there is no global consensus, no policy.
It is totally holding hands and helping out - to Schmidts, Trumps, Musks, Epsteins. Just not to poorer people.
> Do you think AI will go away and suddenly businesses will start hiring people back?
In fact, with well run economy that systematically prevents monopolies, yes it tends to hire people no matter what technological level. Currents state where few super powerful companies are able to push themselves into everything and create monopolies via dumping prices, even as they are not profitable and can count on their friends in administration to bail them out once if all goes pop is the ineffective economy.
I agree with most of your points except the last one.
Probably easier to ask a question than argue a point:
How eager are you to use Government services?
Businesses are only well-run if they make profit: hiring the cheapest labor to produce something people will actually spend money on. And the more frictionless that process, the more our economy advances it. AI fits in there very well.
I also want to point out: startups are usually happy to be bought up by the bigger guys.
The economy looked really good before the dotcom crash too. The crash didn’t make the internet go away, but it damped the hype and blind, unchecked optimism that was leading to some rather short-sighted decisions.
Right now so many companies are trying to use AI just to use AI, rather than using it when and where it actually makes sense. This is the big thing that drives me, and I think many others, a bit crazy. I don’t expect a bubble pop to make us go back in time to 2022, but I expect it will put an end these the AI mandates, token maxing, and other foolish behavior.
> The crash didn’t make the internet go away, but it damped the hype and blind, unchecked optimism that was leading to some rather short-sighted decisions.
The crash did not make the internet go away. I don't foresee a world where we will go back to the pre-AI times either. In the same way that post dotcom crash, you would be a fool to not have your business online, I think we will find similar things to be the case around AI. Even if the bubble bursts AI is here to stay and that will have major consequences for labor.
The irony of the dotcom crash is that a lot of 'dark fiber' service started rolling out decades later. Fiber that was laid during the dotcom era.
There are lots of datacenters going up in similar fashion. I don't know if they'd have the same utility decades later (very unlikely), but it's interesting.