600K is the amount a redneck can pull working in the oil field in upper management. It is by no means "wealthy" unless you are in the Midwest or working as a white collar paper pusher in a field that doesn't pay well. It is high compared to a Starbucks barista perhaps but go to places like New York and Boston and the Bay Area level of salaries hardly stand out.

You can make 600k a year running small businesses like online tutoring too. Too many people (especially in tech) get overly caught up with hype and passion projects and refuse to put in the effort to remain dynamic and keep up with advances in their industry. If LLMs are the latest thing, go learn how the engineering and science side of language models work before launching yet another side project without doing any market research. Sure it's an empirical field, but you can learn a lot from simply replicating paper results. Drop that React tutorial and learn how variational learning works. Train and fine-tune a model yourself. Don't just blindly call an API and build products on undifferentiated factors like prompt engineering.

The writing's on the wall right now that machine learning engineering is what the industry will be embracing in the next 3 years. It may be partly hype, but it is obvious that there is immense value to be captured.