Flood of Fake Science Forces Multiple Journal Closures | Hacker News Reader

Flood of Fake Science Forces Multiple Journal Closures

wsj.com

33 pointspaulpauper9 comments

We're at a point where the average peer reviewed paper isn't noticeably better or inherently more trustworthy than the average preprint on arXiv/bioRxiv/medRxiv/techRxiv/etc. Peer review adds a lot of cost, delay, and friction to the publishing process, but I don't think that it contributes very much -- the "peer" comments I've received on some of my papers were bafflingly obtuse and unhelpful -- and it could quite fairly be said that companies like Elsevier are openly villianous.

I think that new paradigms are needed, but certainly more preprint servers are always a good thing.

If one must publish in a journal, I have a handy rule of thumb: Only publish in a society journal. If the American Society of X, or the European X Society, won't publish your paper, then throw it up onto a preprint server, save it for a symposium, post about it on your blog/substack, or forget about it. Low-quality pay-to-play journals are shameful.

https://archive.is/5k8nI

To the average person, science must seem less trustworthy than ever.

The average person outside academia doesn't read research papers. They read 3-minute news articles with titles like "Study suggests X is related to Y", which tend to have pretty obvious findings.

> which tend to have pretty obvious findings

Perhaps I'm biased, but instead of "obvius" most scientific news in newspapers tend to be overhyped doubious results.

Because it is less trustworthy (not the method, obviously, but the institutions). The most trustworthy elements seem to have more in common with engineering than science.

Does science as published help us end capitalism? I think it doesn't. I think on the balance, science benefits capitalism and prolongs its systemic destruction of human dignity and the planet.

The knowledge collapse is the AI driven degradation of knowledge production systems. This is caused by a flood of AI junk, which reduces the amount and discoverability of high quality knowledge.

Assuming that science entrenches and prolongs capitalist abuses, then there's an upside to the knowledge collapse: we can use it as a weapon to help collapse capitalist knowledge production loops while making knowledge more accessible to anticapital efforts.

Ways we can contribute:

1. Refuse to build capitalist solutions to knowledge collapse. If you know how to build a non-collapsing RAG system, you can build it for a worker owned commune. But refuse to do it for a capitalist or state organization.

2. Contribute to open source knowledge collapse systems that can help people create fake science, and generally contribute towards knowledge collapse.

3. Contribute to the library Genesis, the world's #1 source of equally available knowledge for all, and one of humanity's most important communal resources

4. Contribute towards open source tools that are good at creating anti capitalist knowledge and organization, but hurt corporations due to their openness and freedom

It's always surprising to run across unusual worldviews that just espouse it as if it is normative. Is the hope that people will be fooled into thinking it is a commonly held belief, and then take some sort of action from that? Or perhaps the commenter is genuinely oblivious to the oddity of their sentiments?

I am partly oblivious to the fact that it should be considered odd, and I'm not trying to peer pressure folks, but provoke and inspire them.

My arguments are there for anyone who wants to consider them. You might choose to ignore them and characterize me as naive or deceptive instead. That doesn't dismiss them.

There's still a major conflict where science, states, and corporations are working together in a system that harms people. Analyzing and opposing that conflict is important. As Scientists and software developers, we have a unique call to take a stand. If you don't feel that call now, maybe you'll feel it in the future!

Good luck with your work, and thanks for your response!

Reply on news.ycombinator.com