Plague of our ages I guess. Ironically AI might even make it worse.
And then we'll wait till the next bubble.
Gains seem to have leveled off tremendously. As far as I can tell folk were saying "Wow, look at this, I can get it to generate code... it does really well at tests, and small well defined tasks"
And a year or a year and a half later we're at like... that + "it's slightly better than it was before!" lol.
So, yea, I dunno, I suspect we'll see a fair amount fall away and some useful things to continue to be used.
Also you seem to forget that irrespective of cash profits in the future, will this investment generate excess returns? Nope. That's what investors care about. Its not even profit actually.
Beneficiaries are the ones who care about the actual tech and what it can do for them. Investors are the ones who care about making money off the tech. For the Beneficiaries, AI hype is about right where it should be, given the demonstrable power of the tech itself. For Investors, it may be a dangerous bubble - but then I myself am a Beneficiary, not an Investor, so I don't care.
I don't care which companies get burned on this, which investors will lose everything - businesses come and gone, but foundational inventions remain. The bubble will burst, and then the second wave of companies will recycle what the first wave left; the tech will continue to be developed and become even more useful.
Or put another way: I don't care which of the contestants wins a tunnel-digging race. I only care about the tunnels being dug.
See e.g. history of rail lines, and arguably many more big infrastructure projects: people who fronted the initial capital did not see much of a return, but the actual infrastructure they left behind as they folded was taken over and built upon by subsequent waves of companies.
My nuanced position is that it's great in some niche scenarios - speech to text as an example, or for isolating instruments in audio - and vastly overhyped in everything else, like LLMs. It's a mediocre google searcher at best.
How so? I am a beneficiary of AI, and I think it is overhyped. Investors overhype it since they are invested and want to create a bubble, since bubbles makes many people rich, the more hype the longer the bubble last and the more money investors make.
No investor want to say they invest in a bubble, as then they would not invest, if they invest they want to surf the hype. So the doubters are not investors, the doubters are AI beneficiaries that just feel the benefits aren't that large.
People seem to like the ability to summarize things quickly, and quickly scaffold up presentations, reports, and flows I guess?
Regardless, whatever it is that the masses decide is worth paying enough to keep the business afloat will probably survive.
And the rest will... dissolve.