Hurts to read. But I guess the scientific publications were already headed that way and AI accelerates mostly the unscrupulous. Now it's going to be harder and harder to separate the wheat from the chaff, and I fear AI will not help here, unless it's about purely theoretic results. I guess a web of trust will be more and more important in the future. Not a great solution though.
You haven't read the good ones it seems. The number of bad papers published is irrelevant for progress. It is just important that there are good ones and that the researchers who can advance the state of the art on top of these good papers can find them in the sea of bad papers.
I fear you need to bootstrap by finding people you trust not to produce slop. Check what they use/build upon. If you do this for multiple independent people in a field, the things used by many should be good.
It's neither quickly done, nor 100% bulletproof, but I guess science never was.
I think this is a problem that existed even before. If you look closely, AI just accelerates everything—both the pros and the cons. It seems to be an inherent problem with the academic system itself.
Signals are inherently bound to be distorted. The "legitimate methods" conceived by humans and social signals are easy to manipulate. Because public data like the number of published papers and diplomas act as strong social signals, they end up being distorted. This is known as Goodhart's Law.
The problem is that if this persists long-term, there will eventually be a shortage of personnel to quantitatively evaluate papers, and society probably won't hire these researchers. There is also no incentive to hire them, as verification rarely leads to direct profit. People tend to want to be the starting point of a discipline; they generally do not like checking other people's work.
As a result, academia will increasingly be validated solely by "which lab, which advising professor, and what peer network certified you." The value of the offline world will rise, simply because digital costs have become too cheap.
Soon, research quality and social status will likely merge again, resembling the Middle Ages. Even if we assume someone has actual research capabilities, if evaluators do not invest enough time in verification, the system will inevitably revolve around the more prestigious labs.
I suspect there will be a flood of research applying AI, but the topics of these studies will likely be quite narrow and localized (it would be nice to have statistics on this). I believe the overall diversity of topics explored across academia will diminish.
For this reason, I think existing networks will only grow stronger.
I feel that any solutions to this problem will go awry. I doubt self-purification is even possible anymore. Established researchers protect the monopoly power of publishers, while those existing publishers generally just want to cut costs. The authority of these publishers must be protected by the established researchers who have already secured their positions.
Next, the system that used to manipulate the number of papers will probably shift into a system that manipulates network reputation. The competition to get into good labs and the strategies to secure co-authorships with famous researchers will simply become the new means of competition.
Modern Academia has largely always rooted very deeply in the self and a sense of ego, and not within progression of the human knowledge base. This is in large part due to an “Us vs Them” mentality that has been brewed over several decades even before the World Wars. The consequences are that individual output and prestige are placed higher than meaningful contribution, and that people do not seek truth but rather seek self-gratification.
In my experience this is more prevalent and can be hidden better in the soft “sciences” like history or psychology, but you can even find it in fields like theoretical physics. In fact, a perfect example lies in the recent solving of the 700+ math proofs via OpenAI. The mathematicians were not upset that an LLM did the work, they were upset that a human could not take the credit. This is because that credit is the currency within academic circles, and OpenAI “deprived” those researchers of putting their name on the paper.
As for the reputation of networks, we already see that taking place and the fraud capable in that, too. A letter of recommendation means little when they can be bought or obtained dishonestly.
Ultimately, I believe that much like in the Middle Ages a Renaissance will form of independent non-Academia scholars who will push the field. I already follow a handful of YouTubers who I respect more than any Ph.D’d pencil-pushers.
For a small list: Nighthawkinlight, Tech Ingredients, Cody’s Lab, and Hyperspace Pirate
Here's where we talk about whether AI is taking our jobs; it took me 5 years and three supervisors to produce a doctoral thesis that nobody will ever read.
My wife actually had her paper cited recently, and that sat dormant for a decade. Hopefully yours does too, it’s quite a funny situation to find yourself in.
Though I’m not quite sure how that ties into AI replacement theory.
I don’t think we’re ready for the absolute warp-speed break down of reliable references this process of slop creation, citation, and publication that is going to occur.
I look to AI as an unprecedented tool to clear out the slop and find the well supported work. But no, the slop pipeline pays its way and finding a way around it is not very important it seems.
> "a Spanish fellow named Álex Escolà-Gascón, Professor for Exorcism and Divine Apparitions, fabricated a bizarre clinical trial to prove the existence of immortal soul and life after death."
There are days where I feel the title "Director of Exorcism and Divine Apparitions" would be appropriate for what I actually do. :-)
Not surprised it reached books. In Graphite's Common Crawl sample, ~50% of new web articles have been mostly AI-written since late 2024, and still ~50% as of May 2026. The odd part is the plateau. Chart: aiwords.si
18 comments
[ 0.22 ms ] story [ 20.1 ms ] threadYou haven't read academic papers in a while...
> can find them in the sea of bad papers
How to find the good ones, as an amateur in a given field?
It's neither quickly done, nor 100% bulletproof, but I guess science never was.
Before it was artisanal non-replicable, p-hacked, publish or perish motivated, politically biased slop. Now we have some AI sugar sprinkled on top.
Signals are inherently bound to be distorted. The "legitimate methods" conceived by humans and social signals are easy to manipulate. Because public data like the number of published papers and diplomas act as strong social signals, they end up being distorted. This is known as Goodhart's Law.
The problem is that if this persists long-term, there will eventually be a shortage of personnel to quantitatively evaluate papers, and society probably won't hire these researchers. There is also no incentive to hire them, as verification rarely leads to direct profit. People tend to want to be the starting point of a discipline; they generally do not like checking other people's work.
As a result, academia will increasingly be validated solely by "which lab, which advising professor, and what peer network certified you." The value of the offline world will rise, simply because digital costs have become too cheap.
Soon, research quality and social status will likely merge again, resembling the Middle Ages. Even if we assume someone has actual research capabilities, if evaluators do not invest enough time in verification, the system will inevitably revolve around the more prestigious labs.
I suspect there will be a flood of research applying AI, but the topics of these studies will likely be quite narrow and localized (it would be nice to have statistics on this). I believe the overall diversity of topics explored across academia will diminish.
For this reason, I think existing networks will only grow stronger.
I feel that any solutions to this problem will go awry. I doubt self-purification is even possible anymore. Established researchers protect the monopoly power of publishers, while those existing publishers generally just want to cut costs. The authority of these publishers must be protected by the established researchers who have already secured their positions.
Next, the system that used to manipulate the number of papers will probably shift into a system that manipulates network reputation. The competition to get into good labs and the strategies to secure co-authorships with famous researchers will simply become the new means of competition.
In my experience this is more prevalent and can be hidden better in the soft “sciences” like history or psychology, but you can even find it in fields like theoretical physics. In fact, a perfect example lies in the recent solving of the 700+ math proofs via OpenAI. The mathematicians were not upset that an LLM did the work, they were upset that a human could not take the credit. This is because that credit is the currency within academic circles, and OpenAI “deprived” those researchers of putting their name on the paper.
As for the reputation of networks, we already see that taking place and the fraud capable in that, too. A letter of recommendation means little when they can be bought or obtained dishonestly.
Ultimately, I believe that much like in the Middle Ages a Renaissance will form of independent non-Academia scholars who will push the field. I already follow a handful of YouTubers who I respect more than any Ph.D’d pencil-pushers.
For a small list: Nighthawkinlight, Tech Ingredients, Cody’s Lab, and Hyperspace Pirate
Though I’m not quite sure how that ties into AI replacement theory.
There are days where I feel the title "Director of Exorcism and Divine Apparitions" would be appropriate for what I actually do. :-)