Perhaps I'm biased, but Amarita Natt, Ph.D. is spot-on here:

Interesting. I've been saying AI is a bubble for quite some time and have also compared it to the dot com bubble. I think when the dust clears AI will still be present but that the use cases will be more realistic, more solid, and arguably rather "boring". That's not a bad thing, by the way - databases are by and large incredibly unsexy and also critical to literally all of our modern life. I'll be interested to see if Altman's prediction is accurate.

Not only will future genAI use cases be boring, but:

1/ There will be fewer use cases – because not everything calls for genAI. We'll learn what works, and we'll stop throwing spaghetti at the wall in search of the next hot thing.

2/ Even use cases that are exciting today will become boring and/or invisible in the future – because that's what happens with technology. We like it, then we take it for granted.