That's funny, because to me these headlines about how deep learning is over-hyped and hitting the wall remind me of headlines from ten years ago about how... deep learning is over-hyped and hitting the wall.
And yes of course hallucinations are a huge problem for most of these use cases, but they aren't stopping people from using them anyway. We have a new misinformation problem and it has no agenda. It's basically just white noise.
So my money is also on this changing the world dramatically, just not in the in uniformly positive way that the hype said it will.