You implicitly assume, LLMs are actually important enough to make a difference on the geopolitical level.
So far, I haven't seen any indication that this is the case. And I'd say, hyped up speculations by people financially incentivized to hype AI should be taken with an entire mine full of salt.
It's an economic benefit. It's not a panacea but it does make some tasks much cheaper.
On the other hand if the economic benefit isn't shared across the whole of society it will become a destabilising factor and hence reduce the overall economic benefit it might have otherwise borne.
They seem popular enough that they could be leveraged to influence opinion and twist perception, as has been done with social media.
Or, as is already being done, use them to influence opinion and twist perception within tools and services that people already use, such as social media.
I think ground-zero for that line of thought is with Leopold Aschenbrenner[0]. Who I believe now runs an AI focused hedge fund.
The same stack is required for other AI stuff like diffusion models as well.
First, its not just about LLMs. Its not an LLM that replaced human drivers in Waymo cars.
Second, how could AI not be the deciding geopolitical factor of the future? You expect progress to stop and AI not to achieve and surpass human intelligence?