Is AI hitting a wall?

Coinsreporter

I think that it would be good if there's a plateau in the current capabilities, because:

1. It will force radical redesign over the long term, which is great because the current tech does not live up to the promise, and as the article notes: likely won't in the future either.
2. It will allow for consolidation, where open models can commoditize the capabilities that are currently reserved for expensive hosted closed models
3. It will allow for time to optimize what we have today, require smaller hardware for runtime and make it sovereignly accessible to the masses on affordable devices (this is also said to be China's long-run play because they want to manufacture these devices)