That's not exactly what the main problem is. AI is getting better and better and in some fields, it's already better than the average human.
But herein actually lies the problem. AI is becoming so good that many just accept it as being the truth without really looking at it. So people let AI solve their problem. So they solve their problem faster than before, but don't get the experience of looking into it, failing, getting grip on the problem and then solve it themselves. This means, over time, experience and knowledge get lost. I already said that AI is just restructuring things it already knows, but it's not really innovating. So it's clueless when encountering something really new. But in due time there won't be people with the insights that people used to have, having gotten their hands dirty. This means progress will stall.
So my prediction for the future is that AI will initially accelerate innovation, but it will reach a plateau once the experienced people are gone. This will be sooner than you think. I've already seen students who are clueless without the use of copilot. And this is only, say, 3-4 years into the hype?
