I obviously don't have the data, but I would argue that a tiny tiny fraction of Google's energy usage is dedicated to training language models. I'm guessing th

I obviously don't have the data, but I would argue that a tiny tiny fraction of Google's energy usage is dedicated to training language models. I'm guessing the vast majority of the energy usage is for serving billions of requests for various products, such as search, youtube, maps, gmail, photos, etc.., and the cpu, network, and storage requirements for those requests. Training and retraining ML models is definitely not on the hot path.