Comment by alfalfasprout

Right. Which means tokens are actually being priced well under cost once you factor in all this datacenter/GPU capex. Also worth noting the datacenters are not purely for training. They're for inference too.