Nicholas Polson has authored 258 academic papers in 2026 so far
statmodeling.stat.columbia.edu
statmodeling.stat.columbia.edu
If one were inclined, they could shorten the title to "AI Synthesis Across Economics, Psychology, Biology, Philosophy, Game Theory, and Spiritual Tradition." Maybe that's an Easter egg.
AI submissions, soon AI reviewers, and shortly after only AI readers.
The guy's name is "nosloP" spelled backwards. Maybe spelled normally it's supposed to signify the opposite.
Most insightful comment
This is already happening. Write code with an AI agent, agent creates PR, other agents review, coding agent iterates, PR gets merged, and hopefully real humans use the application and provide feedback
The latter step is also now staring to be replaced by AI (user directs agent to access app, get data, perform actions)
Similar thing with email
It isn't even partially human written, it's just an LLM's output.
Oddly it seems to have happened on longer posts? Don’t know if that may play a part.
I mean it's everywhere now. I complained already at my workplace multiple times, we get huge Claude generated documents, which people then ask Claude to summarize. I don't mind some of it in principle but people are too lazy now to input all the details so you get a bunch of fantasies that were explicitly ruled out already. It's tiring.
How do i know? Because i doubt "Access&Identity will help us push forward the productivity of all collaborators" was what the team responsible for the prompt wanted to say.
Reverse nominative determinism
For some weird reason in the last few years people forgot the differences and now any PDF is an "academic paper" [3]. (I'd be more annoyed, if the internal system of the university also forgot the difference. I'm afraid to ask.)
[1] Whatever "serious" means, that's a huge can of worms.
[2] They get an doi and a version number in case of a change.
[3] A few years ago they were call "whitepapers".
Source: https://statmodeling.stat.columbia.edu/2026/08/27/258/#comme...
Sometimes I just write an email (in text) but sometimes it's a short LaTeX document (3 pages, title, date, my name, may have an introduction to ensure we use the same notation, usually no references because I'm lazy). Some of my coworkers are using AI that likes to make longer stuff and format it like a letter to the Nobel foundation.
Some are good, most are wrong or bad. Some are reproductions of others research work, some are early experiments, some are rediscovering the wheel, some are misguided ideas, and lucky a tiny part are good. Anyway, they are for internal use, we don't try to contribute to the thermal death of the universe uploading that crap to a free preprint server.
I don't understand:
a) Why they upload internal notes/reports to a preprint server?
b) Why are these reported as "academic papers"?
You add up all the claimed activities and then realize it’s more hours than are in a day so someone’s lying.
I don’t see how this author can claim that these papers didn’t “…suddenly materialize from a magical “chatbot” in 2026” when it’s obvious that 1-2 people are not producing 258 papers that are tens of pages each in 9 months with any sort of academic rigor.
What was even more hilarious was the fact that he had visibly no idea of what the game was even about (he was bragging about item level instead of stat and dismissing the best item of his gear because it was lower iLevel than the rest).
If the papers were of a uniformly high standard then I might not have a problem with it.
If they are hoping some of them are good and the review process will filter out the bad then it seems like they are not doing research, they are writing hypotheses and at pushing the work of testing them onto the reviewers.
I could also see that being done to make some sort of statement about the review process, but it's not clear if that is their goal.
He's wasting people's time, which is already the resource that many people lack any to spare. I'm not sure if he realises this part.
When you compare that to the consequence of the hydroxycholoquine paper, which is comparable in public policy impact, where Didiee Raoult lost all its legitimacy and his job, it's easy to see that economics as an academic field doesn't value probity much.
Oh also at least 40% generated peer-reviews are contributing to this statistic.
"Show me the incentives and I will show you the outcome"
If moral, they could have decided that propagating their message will improve the world, even if only at x% efficiency due to use of AI, and that their intentions will take precedence over being accused of slopping.
If immoral, they could have chosen to run a grift on the chance that no one would notice, overdone it, and then played every card in their hand while perhaps sweeping away evidence of their slop once caught.
(These are constructed teaching examples, not situation-specific inferences or claims.)
For condensed knowledge on highly specific topics (i.e. more specific than is economical in book form) I often know nothing better.
Most recently I did sort of a self-taught crash course in ground-penetrating radar and applications for an archeological endeavour I write software for, and a number of papers have been really invaluable.
I believe this is one of the LLMs' more destructive elements. The more effective ones are quite good at providing output that gets the meat to repeat the output to other meat. Whether the output actually benefits the meat is disconnected from whether or not the meat accepts and repeats it.
That is not realistic, but I suppose where things are heading is that you have some indicator of the strength of evidence -- see Fig. 13 in the following insightful take:
https://news.ycombinator.com/item?id=49407226
Though, "strength" should probably be "reliability" and "validity", and I suppose those indicators are more for picking signals from the noise; i.e., what is even worth clicking and reading. That would be increasingly valuable already today due to the volume (and, yes, slop and other related stuff).
And to expand on this, it's not realistic because science is not armchair philosophy. You have to go out and measure the world.
Sometimes, through force of will, a person can think deeply about a problem and come up with beautiful theories that explain our measurements. Many scientists had careers like this, probably most famously, Einstein. But it's worth noting that Einstein also got a lot wrong! [1]
Even if we somehow give an LLM the ability to go out and measure things, I seriously doubt that the role of humans in science is done. There's a big difference between "an explanation" and "a good explanation." Ask any physicist. There's a surprising amount of aesthetics involved. Good theories are consistent with the evidence, but it's more than that-- there's a great deal of "taste" involved. And there's a good reason for that. For any real problem, there are effectively an infinite number of alternative hypotheses. From a "theory of science" standpoint, this should cause scientists nightmares, but it doesn't. Because by the time you are a practicing scientist, you've developed a feel for what constitutes a satisfying explanation. If you spend time with scientists, especially in the "hallway track" at a conference, "taste" is a frequent topic of conversation!
[1] https://en.wikipedia.org/wiki/Einstein%27s_unsuccessful_inve...
I work full time on "lab in the loop" AI, so I'm pretty familiar with the need for real-world experiments. I am not proposing a fully autonomous scientist that could read an arbitrary paper and emit whether it's universally true without some verification method.
Also, to your statement: " Because by the time you are a practicing scientist, you've developed a feel for what constitutes a satisfying explanation."
I'm a practicing scientist (well, ex-scientist) and it seems like most "satisfying explanations" end up being wrong or incomplete simply because they seem so satisfying.
What's the difference? How do you find real mistakes without a model? Either you have a trusted mathematical model (in which case you already have a complete explanation) or you have to compare it against the ultimate oracle: the world. Or are you proposing something like "let's use an LLM to convert this hand-wavy English paper into a formal proof and then check it for logical fallacies?" In which case, fine, that would be useful, but that's not exactly the same thing (and also not as important) as saying that a paper advances a bad explanation. Just that the explanation is flawed in some way.
The next example I can think of- I am not sure it qualifies. I read a paper where they deleted one gene at a time in yeast (it has 6000 genes) and determined whether the mutated yeast could live or not. For each gene where the yeast died, they added that to a list of "essential for life" genes. The paper concluded they had found some interesting proteins that should be studied. I read the paper and the first thing that sprang to mind, are any of these genes overlapping? Because we know (somebody already demonstrated in a lab) that genes do overlap (which is truly weird!)
I wrote a script and showed that every gene they reported as essential for life overlapped an already known gene that was essential for life. I wrote the authors, who never responded, but wrote a followup paper where they acknowledged they probably had a high false positive rate due to overlapping genes with known fatal effects. My guess is you'd say that either I used a model (existing literature) or I compared against the world, but realistically, what I did was trivially come up with a better explanation than the authors. T hat's what I want LLMs to do for me, and it seems like the direction LLMs are going will fulfill my desires.
To me, a single significant error of any kind brings the entire paper into question. If a less important figure contains an image duplication, that makes me wonder if I can trust any of the images.
In any case, my point was that rejecting results due to technical flaws is intellectually lazy. As a scientist, your job is more about trying to find value in other people's work than finding excuses to reject it.
I disagree with your premise. Part of our job as scientists (thankfully no longer mine) is to reduce the irrelevant and incorrect noisy as early as possible. I have seen so many grad students get excited by a paper and put enormous effort into reproducing somethign that was a false or fake result.
If you want to determine reliably whether something is irrelevant and incorrect noise, determining whether there is anything of value is a necessary first step.
I've seen many reproduction attempts in bioinformatics fail, because they person trying to reproduce the work didn't have the conceptual background to do it correctly. Instead of spending enough time studying the theory, they rushed directly to action.
I also left a postdoc position over my professor's decision to rewrite my paper to juice all the stats- not a specific error, but selectively interpreting the data to make the results look better than state of the art, when they were not.
I would absolutely love to have my "paper correctness AI" mark all those bad reproductions in the literature and outright misrepresentations- it's all too easy to rush to publish and get a lot of attention- especially if your advisor or coauthors are prestigious and mildly unethical.
People in the early days used to often whine that LLMs just regurgitate text snippets (unfounded of course), but I think the way we currently train and RLHF them actually seems to largely make them unable to reproduce the knowledge they have been trained on, since they seem to just always want to please the mean with their output. I'm oversimplifying the mechanisms, but you get my drift.
You ask the LLM for papers in this special field that you are interested in.
It does the search, fetches 20 papers it finds relevant
Summarizes them or points to ideas with reference
Polson is just pushing the knob to 11 because it's louder.
We've been in the dark ages of science for so many decades...
Edit - people are assuming that i want this to happen. no, i don't. i am reading more than i ever did before, maybe out of rebellion. i am just stating what I THINK is to come. it might not go this way at all.
He was supposed to present a topic he learned about at the conference. Instead, he gave a cheeky talk which introduced the very concept of written language to the audience, as if we had never heard of it. Starting from the history of writing on tablets and so on, and ending in practical advice like, "when words are written down, you don't have to remember what they were exactly; you can just read the words again." I loved the chutzpah and could barely hold myself from chortling out loudly at some of his slides.
There has always been bad writing, but historically it has been easy enough to filter the garbage out. Now we get garbage that looks so much like real thoughts that it's hard to tell. This is not a positive development. Separating the wheat from the chaff is extra work, and the overhead is a burden on humanity. I suspect that this is a burden that almost nobody actually wants, but we're getting it anyway.
Why should anyone bother to think, if they can ask AI for a well researched, correct, and.. khm... "aligned" answer ?
but i am making the case of in-person chat or public speaking, where in-person and public are the relevant parts.
I really don't like AI for document writing as a whole, especially within organizations. Code can be a form of communication, but for the most part it's an engineered thing that can be re-run, tested, verified and so on. But documents drive decision-making, and decision making is often one-time. Decisions should be studied and understood, and depriving authors of much of both is not having any good effects.
The fact that most teachers refuse to do in-person oral exams has always been due to laziness, incompetence, and the reality of the economics of dealing with classroom sizes where 30 is considered small.
Your intelligence and competencey are going to be 1:1 correlated to the quality and quantity of tokens a billionaire gives you access too (local, sure, they control the hardware markets as well).
Keep excercising your writing and reading muscles because it will definatley set you apart from the captive masses in the future.
I wouldn't blame "AI" for that, TV and Youtube and Tiktok did a good enough job.
Perhaps this will help bring more scrutiny to the metric of publishing quantity not being as valuable as it has been treated.
Journals get less money from less submissions, conferences get less money from less submissions, manual reviewing takes more time which decreases the time you can spend generating papers yourself, less submissions means less acceptance means less funding grants.
The only possible thing is some idealistic demogogue take that knowledge will get diluted, it will be bad for humanity, etc, which moves the needle exactly 0 in a capitalist society with our incentives.
If you only knew.
> As someone who has known Nick in the past, I’m a bit surprised by this multitude of articles and books. However, I think you’ve missed a fifth possibility that would suggest gentle handling: a mental health issue. The papers and books on philosophy, religion, and issues well beyond anything econometric plus the videos on the Riemann Hypothesis suggest someone who is no longer bound by the intellectual boundaries or social mores of proper research.
I don't know whether the commenter is legit, and whether this actually happened, but I wonder if AI-psychosis (or at least, AI-amplified psychosis) might not become a real thing. People already on the edge having "productive" conversations with LLM chatbots and going down rabbit holes. It can't be a good thing, especially since LLMs are sycophants that indulge every crazy idea you have.
I'm not saying this didn't happen before, I just wonder whether LLMs enable a whole new scale of crazy.
Can’t imagine the utter garbage that Claude will produce when you give it a simple prompt and let it write thousand of pages.
Kind of makes me wonder about all that AGI talk if these models can’t even the work of a good PhD or postdoctoral student.
0. https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=170...
1. https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=228...
2. https://www.chicagobooth.edu/faculty/directory/p/nicholas-po...
The Everest effect.
George Mallory: "The first question you will ask, and the one I have to answer is, 'What's the point of climbing Mount Everest?' And my answer should be, 'It's no use.' There is not the slightest prospect of any gain. Oh, maybe we can learn a little about the behavior of the human body at height, and medicine can change our observations for aviation purposes. But nothing else could come of it. We will not bring back a little gold or silver or gemstones or coal or iron. We will not find soil or earth that we can grow crops that produce food. There is no point. So if you can't understand that there's something in humans that responds to the challenges of this mountain and goes to meet those challenges, then the struggle that ensues is a struggle from life itself upwards, and you won't see why we're leaving. What we get from this adventure is just joy. We don't live to eat and make money. We eat and make money to enjoy life. That is the meaning of life and the purpose of living."
Why generate all of these papers? Because, for the first time, he can.
Doesn't have to be for the money or fame or potential for speaking engagements. It could simply be that he wants to test the limits of what's possible.
The dopamine hit of having even their most speculative ideas expanded in a convincing paper format is self-reinforcing and soon the affected individual starts believing that the output is perfect and needs no review, because reviewing it would delay the next stimulation cycle.
Correctness, formal verification, using your own cognitive resources, deterministic results are in many ways actively counter productive to the kind of effect people want to get out of this
What a disgusting, tasteless joke.
it was right in front of you :)
Sin: A Slop Colon!
(For instance, 58 pages, 80 pages ... I don't think of these as normal academic length if this is the main body content ...)
[1] https://docs.oracle.com/javase/specs/jls/se17/html/index.htm...