That’s all correct, of course, but it represents somewhat of an ideal case.

First, yes, you have to secure funding. Your best chance is to be someone who has already successfully completed grant-supported research with a solid history in your field, and to be looking at something considered sexy at the time. I’m not in the business anymore, but I shudder to think about how many grant proposals are offering to use LLMs. The rub is that grants can be tough to get - there’s orgs within the NIH that have a less than 5% acceptance rate. Let’s say you’re lucky and you get your grant. Depending on your institution, a big chunk of that goes into administration. The rest is for you and your colleagues and students and lab workers and so on, as well as equipment and other expenses.

You also will probably want to hold back about $10k or more for publication fees. Many journals do not require a fee to publish, but do require one to make the paper open access so that others can read it without paying a $30 fee for a single paper. When I was doing it the fees were usually between $2-3k per paper. It’s not that big of a deal if your grant is $500k, but it can be quite a chunk of money for smaller grants. In any case, you’re paying someone to print your paper, which you wrote and edited and which was reviewed and recommended for publication by other unpaid academics. If you cannot pay the fees, your work will not be accepted by most open access journals, and will not be open access if accepted by a paywalled journal.

It is not true (at least in the US at the time I was doing it) that government sponsored research will be open access by law after peer review. We fought hard for it, but the publishing lobby is pretty strong. I think the law is currently that government financed research must be made open access within a year or so of publication.

The problem comes with smaller institutions and less well known researchers. I had a friend who was a professor of finance at a smaller university, and he had to pay out of pocket for his publications as well as some of his conferences. And their salaries aren’t that high in any case. He had hard money - his salary for teaching classes - but also had to keep publishing to keep his job and advance. I had another situation where I was publishing a paper in a very small but within its subject prestigious journal, where I was more than happy to pay the pub fee. The editor told me quite frankly that he was working with a researcher from another country who was trying to figure out how he could afford to pay the pub fee because he said our paper would essentially be paying for his as well.

So, after all of that, I do consider the academic publishing business predatory and parasitic. Here’s how to get papers for free - legally. I’m not touching on any other means.

1. Search for the title - in quotes - that you’re looking for. You can find individual papers by their abstracts, which generally are made publicly available. There are preprint services like ArXiv where researchers upload their papers before they’re published. If it exists, most of the time a published paper will have its final form available as a preprint with the layout being the only thing that changes. It makes sense to check though.
2. Go to the author’s website. Researchers will often have links to their publications on their professional page.
3. Write to the researcher and request a copy. We love that. You might need to ping them a couple of times because people get busy and forget things, but overall you’ll probably find someone who would be very happy to send you a copy.