Hacker News Reader

Comment by augment_me | Hacker News Reader

As a PhD student, I am seeing the same thing. I am currently generating about 10 papers per month for marginal increments/architectural additions that are autoreviewed in a loop by Claude + Codex until those tools don't spot any major issues and then sending them in to various conferences. Since more than half of the reviewers are just using these tools to review, it's really easy to get a good score. This year I got into ICLR, ICML, NeurIps, SIGGRAPH and ACL with 2-authors papers (me and my supervisor).

Its fucked up, but I am getting a lot of praise from the department/supervisor, I got an very good internship lined up next summer, and I have plenty of time to study whatever I find meaningful because the slop generates itself. I think that this ship will sink if enough people do this, so I set up similar systems for some of my colleagues as well. The only way to break the system is to overload it until it doesn't work, so I can only hope for this.

View on HN

>The only way to break the system is to overload it until it doesn't work, so I can only hope for this.

This is a joke, right?

Not really, with a slow transition you will just have people in power clutching on to their careers, tenure and more general gatekeeping with whatever means they have. You can already see this starting with the paper count limitation that is 1 for new scientists and 20 for established scientists for ICLR 2027.

I have in fact set up similar autosubmission systems for my peers who wanted it and they are already finding success with it 2 months later.

I’m reasonably sure this person is either lying or at least misrepresenting their success. I did some vibe-coded searching through DBLP and there are very few people who have published in all of those conferences without a publication prior to ‘23 (an arbitrary cutoff to rule out big name researchers with labs), let alone with only 2 authors. ICLR only just closed entries, but I included 2025.

But people are clearly attempting to do this, and some are going to get successful. Putting your paper into ChatGPT prior to submission to get free feedback is really table stakes these days, and you can be sure reviewers are breaking the rules by using LLMs.

Hold on a minute. What's your vision for what will happen once you collapse the system? Revolutions tend not to have happy endings.

I don't really care, I got my ticket out already in case I don't like whatever is coming next.

Best-case, funding will be given with less strings attached, and academia will be about science and discovery instead of status chasing, hierarchy building and number of papers as an incentive.

Worst-case, everything will be clout/social-based instead. Your papers will only be accepted if your dept/supervisor/company has clout.

Either way, the outcome that things will break is inevitable, and the monkey-patching the conferences are applying currently due to the overload like max submission counts and prompt-injections are temporary measures that will not fix the core of that the incentive of paper count and H-index does not make sense in a world where information can be generated for free.

Reply on news.ycombinator.com