It's about product specification. There's no way it fits into their ecosystem without compromising it. Everyone else has compromised theirs and their customers. Apple do not want to do this. I suspect they will fail on AI but will win in the long run as competitors screw their customers over.
I suspect they'll win on AI by not investing $200Bn in CapEx and waiting 5 years until good enough models can be trained for <$1Bn and then just doing that - blocking all other AI models on their hardware beside their own (which their users seem to like) - and saving $200Bn and getting an experience that's good enough for almost everyone and spending enough on marketing to convince everyone it's 10x better.
Apple has BY FAR the best chips at this level. I don't see anything changing that in less than 5 years. That will be a pretty big advantage in the near future.
That's feasible but I think the whole thing will collapse in a heap when the obviously iffy funding structures fall apart due to a mix of lack of declinging investor interest and drunk assery by the US gov resulting in distrust and higher risk profiling of the associated investments.
At least Apple has other products that make money.
It's about product specification. There's no way it fits into their ecosystem without compromising it. Everyone else has compromised theirs and their customers. Apple do not want to do this. I suspect they will fail on AI but will win in the long run as competitors screw their customers over.