Comment by tuwtuwtuwtuw

Why should all AI be home run? I play around with gemma/llama/sd locally but being able to pay some company to do it is very convinient.

I think most companies and people don't want to buy the hardware required to run an LLM like the ones OpenAI hosts.