It is quite astonishing to me how much AI adoption is being driven by literal FOMO.

Devs don't want to miss the boat, lest they become obsolete. Not really doing it for you? That's a skill issue, try harder.

Managers/execs pushing AI through forced mandates, lest they fall behind the competition. Not getting the expected results? It's ok, we're learning. Push harder, set token quotas, do internal hackatons.

Policy makers and investors chasing the next growth engine. Do whatever it takes to stimulate their own local AI industry. Capture the market, returns will come later.

Tomorrow's St Patrick's day, but we've been chasing that pot of gold at the end of the rainbow for a while now.

@plexus This is absolutely spot-on. If you compare this with eg. the bitcoin craze, that was easier to ignore because this was just some new technical concept that could be applied to domains you might not be interested (invested) in. With the AI craze, FOMO is much bigger because developers (and managers, too) and what they are doing daily is at the very core of what the hype is about and it comes with a very threatening narrative ("you will be replaced, resistance is futile").

Also, I have the impression that investors are desperately looking for the next growth thing because a lot of traditional businesses are struggling. At the same time, all indexes have seen tremendous growth over the last years, but apparently greed is always growing quicker than cash. I just wonder what will happen when the AI hype cools down ("AI winter is coming"), which inevitably will happen (sooner or later).

@schaueho @plexus Strong agree that the deeply invasive change to workflow aspect is central.

I think the replacement narrative is stranger. We know LLMs will be used as a scapegoat for layoffs that would happen anyway. It's unclear to me what Claude-like mechanization of the production of code will do to the shape of industry, though. It seems more like spreadsheet software for accountants than google maps for London cab drivers — it is reshaping expectations of the work, but the actually-sensible changes to staffing, funding, and organization seem far more *weird* than "fewer programmers".