[mudrockbestgirl]

You essentially have two choices:

- Get your PhD and try to become a researcher at one of the labs. Super crowded now, but basically the only way if you want to do research.

- Become a ML engineer, data scientist, etc. While you will be working with ML models and understand them to some extent, you may find that it's not exactly what you wanted. You'll likely end up spending most of your time on the same old engineering stuff: Calling into black-box APIs, data engineering, iterating on hyperparameters, building experiment pipelines, dealing with cloud scaling and setting up GPUs, etc.

For example, if you take a look at GPT, 99% of the work that has gone into it is standard engineering/scaling work, not AI-specific work.

[Der_Einzige]

No, you can definitely publish without a Ph.D. You don't have to work in a research lab.

It's as simple as finding a conference, finding the submission style, writing the paper, submitting, and waiting and praying to the gods of peer review.

Of course, you need to have a good idea, but the fruit in AI is still very low hanging.

[mudrockbestgirl]

I didn't say you cannot publish without a PhD. Rather, it's incredibly difficult to get a position in a good research lab without a PhD due to market forces and competition. Publishing in top venues also requires a certain skillset that's difficult to self-study... and these days it often requires money for running experiments. And if you don't want to get position in a research lab / academia, why would you want to publish at all? Getting those citations for your job is the whole point. Otherwise you can just write blog posts.

[la_fayette]

What is the point of publishing research papers, if not beeing a phd student or professor?

Many ground breaking ideas come from underpaid phd students anyway, e.g., yolo.