More than that. At the current discount prices, LLM providers bleed money like mad. They will run out of cash soon if they don't raise the prices to profitable levels. Just double won't cover it.

Which will lead to most customers dropping the LLM services stat. Which in turn will force the LLM providers to up their prices even more...

The last 50 years has shown that the new model is to get a sector of the economy or consumers dependent on a service then once the competition has faded away, jack up the price. Competition either never shows up or collaborates on pricing.

Depends entirely on whether or not Moore's law has hit a physical limit yet.

If we're still doubling computational every X years for the same cost, we'd be able to see same models cheaper. If we're however finally hitting physical limits, then yeah. It won't be cheaper and we can't just smash more context in it like there's no tomorrow.

Then the most likely breakthrough would have to be DNA storage (tri point storage, rather than bi point storage) and then we'd need fast read/write to DNA storage. As that'd theoretically allow for more context for the LLMs, context is the biggest bottleneck

This is if Chinese isn't capable of replacing graphic cards.