Seems fine if LLM capacities are sort of flattening out - there are a ton of use cases just in mass deployment of stuff we already know. But then presumably all the debt fueled AI cap ex is less necessary?
4,49K