Yes, I think you're right that it is not the end of the road for LLM and the application of LLM might be adopted over time across a variety of industries.

I guess what I failed to convey in my original comment was that, like the Internet 20 years ago, the current advancement made by AI might stall at a foundational level, while the landscape evolves.

Essentially, I believe what you're saying is really close in spirit to what I'm saying.