The Future of LLMs

With a domain specific LLM, an operator could actually make intelligent decisions. You can train it as it's starting to learn how to improve deployments. And then the nice thing is unlike a human at the end of the day, might be frazzled and say, gee, I don't know how to did all that. But you can take that model, copy it and put it in another cluster and have same success.