You're not technically wrong, but that looks like tech that might be available decades from now. Even quantum computers might be decades away from practical applications, even though we've been trying to produce prototypes for decades now. Light based (optical) computers are in the same place or probably even further behind.

They'll probably all help in the future, but the current LLM craze doesn't seem to be helped by them at this moment in time, and an economic cycle like this has a boom phase of at most 10 years. So any changes in basic computing hardware will probably help with the next LLM++ tech.