AI Slop Worsens the Degenerative Spiral of Biology Research

 My long 2021 post "'The Degenerative Spiral of 'Grand Explanation' Academia" was one that painted a troubling portrait of malfunction and deceit in the world of academic researchers purporting to have grand explanations for great mysteries of nature. Has there been improvement since that time? To the contrary, things seems to be getting worse. As Exhibit A to back up the claim that the generative spiral of "grand explanation" academia is worsening, I may offer a recent article by Ross Andersen in The Atlantic, one entitled, "Science Is Drowning in AI Slop." We read this, referring to the "large-language models" used by so-called artificial intelligence or AI:

"Almost immediately after large language models went mainstream, manuscripts started pouring into [scientific] journal inboxes in unprecedented numbers. Some portion of this effect can be chalked up to AI’s ability to juice productivity, especially among non-English-speaking scientists who need help presenting their research. But ChatGPT and its ilk are also being used to give fraudulent or shoddy work a new veneer of plausibility, according to Mandy Hill, the managing director of academic publishing at Cambridge University Press & Assessment. That makes the task of sorting wheat from chaff much more time-consuming for editors and referees, and also more technically difficult."

We read about companies called paper mills that help scientists produce papers, either by generating from scratch a fake paper, or by generating dubious or fake paragraphs or dubious or fake images that a scientist can use in his paper. AI makes it much easier for such science fakery to occur. These paper mill companies don't have advertisements with phrases such as "We'll Help You Fake Things!" Instead, they claim to offer "editorial services" or maybe "creative consulting" or "literary facilitation" or some other euphemism.

It seems that many scientists use such "paper mills" to speed up the job of producing a scientific paper, in a shady way similar to a football player taking banned steroids to boost his performance. But with all the AI tools out there, such as ChatGPT, it seems that a scientist does not even have to become involved with external paper mills. Similar results can be achieved from a scientist's desktop, when he uses some AI tool such as ChatGPT. 

We read that AI tools are also being used to write peer reviews. The article says, "Pangram Labs recently analyzed thousands of peer reviews that were submitted to ICLR, and found that more than half of them were written with help from an LLM, and about a fifth of them were wholly AI-generated."  We read this: "AI science slop has spread beyond the journals now, and is also overrunning other venues for disseminating research."

Preprint servers are sites such as the Cornell physics paper server and the biology preprint server Biorxiv.  Andersen states this:

"But in the months after ChatGPT was released, preprint servers experienced the same spike in submissions that journals did. Ginsparg, who is now a professor of information science at Cornell, told me he hoped that this would be a short-lived trend, but the rate of submissions continues to rise. .. A similar influx of AI-assisted submissions has hit bioRxiv and medRxiv, the preprint servers for biology and medicine. Richard Sever, the chief science and strategy officer at the nonprofit organization that runs them, told me that in 2024 and 2025, he saw examples of researchers who had never once submitted a paper sending in 50 in a year."

Andersen ends on this gloomy note, describing the most severe degenerative spiral of the scientific literature:

"When I called A. J. Boston, a professor at Murray State University who has written about this issue, he asked me if I’d heard of the dead-internet conspiracy theory. Its adherents believe that on social media and in other online spaces, only a few real people create posts, comments, and images. The rest are generated and amplified by competing networks of bots. Boston said that in the worst-case scenario, the scientific literature might come to look something like that. AIs would write most papers, and review most of them, too. This empty back-and-forth would be used to train newer AI models. Fraudulent images and phantom citations would embed themselves deeper and deeper in our systems of knowledge. They’d become a permanent epistemological pollution that could never be filtered out." 

These Smart Guys Have Silly Thoughts About AI

For eight years Al Gore was vice president of the United States. In the year 2000 he won the popular vote in the US presidential election, and should have been elected president of the United States. But due to the defects of the US election system, which allows someone with fewer votes to become president, a different candidate became US president. After his election defeat, Gore did long years of important praiseworthy work alerting the public to the dangers of global warming. For this he was awarded the Nobel Peace Prize in 2007. 

Now Al Gore is the chairman of some investment group called Generation Investment Management. Gore recently offered his opinion on so-called artificial intelligence. We read of his opinion in the article here. Gore states the nonsensical opinion that AI systems have a sense of self. He states, "I think that my answer is yes, they have developed a sense of self, in my opinion, that is difficult to distinguish from consciousness.”

In the article we read that Gore makes this feeble attempt at justifying his opinion:

"But as he explained later in this half-hour session, he came to this view by a different path. Gore cited Nobel Prize-winning research by the Belgian physical chemist Ilya Prigogine into self-organizing systems as a model for eyeing how AI models can grow in unexpected ways." 

As an attempt to justify the nonsensical claim that AI systems are self-conscious, this is laughable. The named person (Prigogine) did not do  work having any real relevance to whether artificial intelligence can be self-conscious. His work made claims about physical "self-organization" in mindless, lifeless chemistry or in biological systems, which has nothing to do with whether machines can be conscious.  An examination of lya Prigogine's main work Order Out of Chaos: Man's New Dialogue With Nature shows a thinker who has many a deep-sounding thought about science-related topics, but someone who is not a scholar of minds, brains or computer technology. The book makes no references to computers, except for a few passing mentions of computer simulations. 

 Gore is playing the game here of obscure authority name-dropping. It works like this: you mention the writings of some obscure thinker with esoteric writings on some deep topics, and cite that as your justification for your dumb opinion on some unrelated topic. So, for example, you might say, "I didn't used to think that there were an infinite number of quantum ghost copies of me, but now I believe in such a thing now that I've read Wolfgang von Pauli's work on quantum entanglement." Or maybe you may stupidly say, "After reading Wittgenstein's Tractatus Logico-Philosphicus, I am now convinced the self is an illusion." 

The Cambridge Dictionary defines intelligence as "the ability to learn, understand, and make judgments or have opinions that are based on reason." There is no such thing as real artificial intelligence, because computers don't understand anything. Understanding is something that can only occur within a mind, and computer systems do not have minds. 

 The term "artificial intelligence" is a phony term used in the computing industry to describe sophisticated systems using computer programming, databases and data processing. Computers can do very many kinds of computing and data processing, but no computer understands anything. The fanciest metal computer has no more understanding of anything than a rock in someone's back yard. 

I can describe what gradually happened between 1950 and 2026. The term "artificial intelligence" started out as a purely speculative term, rather like the term "interstellar travel." Just as there were all kinds of speculations and theories about how to one distant day achieve interstellar travel, there were around 1960 all kinds of speculations and theories about how to one distant day achieve artificial intelligence. During one long period, various people released products and systems that were called artificial intelligence programs, but no real effort was made to claim that  artificial intelligence had been achieved.  People were mainly implying that their product (perhaps marketed with literature mentioning artificial intelligence) might be useful in moving towards artificial intelligence. Then gradually companies realized that the phrase "artificial intelligence" was extremely useful in marketing software products. Lured by financial incentives, more and more companies started calling their products "artificial intelligence systems."  It was a runaway snowball effect of hype and misrepresentation. No one had developed any real artificial intelligence, but it gradually became true that hundreds of companies were calling their product "artificial intelligence systems." 

There is still no real prospect of anyone ever developing a computer system with anything like human intelligence.  But what about all those brilliant answers you get from using systems such as ChatGPT, described everywhere as an artificial intelligence system? The output of such a program does not mean computers are understanding anything. What is going on is a clever combination of a variety of things, with most of it being the presentation of text grabbed from web pages written by humans. 

I describe how some of these systems can work in my post here, entitled "What's Called Artificial Intelligence Is Really Just Computer Programming and Data Processing." What is going on is a skillful leveraging of powerful information repositories and powerful technologies such as relational database systems. Here were some of the resources that grew in strength and power between 1995 and 2026:

(1) There arose an internet with billions of web pages, containing many millions of answers to very many millions of questions, the answers being written by humans. 

(2) Almost every book and magazine and newspaper article ever written became stored in some internet location or another. 

(3) There arose enormously powerful web crawlers that could traverse all of these pages, and look for facts and quotes and snippets and answers to questions, that could be stored in powerful database systems capable of combining data in many novel ways. 

(4) There arose countless software utilities capable of performing all kinds of little tasks such as generating a story given a prompt or generating an image given a prompt. 

So-called artificial intelligence systems such as ChatGPT skillfully utilize these resources, combining them with much specialized software.  I don't understand the details of how it all works, but I can tell you something that will help you realize how little novel thinking is involved. 90% of the answers that you will get from a system such as ChatGPT are produced by nothing but a simple retrieval of stored answers. Then probably another 5% of the answers are produced by a simple retrieval of stored answers, combined with a small amount of post-processing added.  Such post-processing is easily accomplished by computer programming and data processing.

Imagine some gigantic building that has 60 floors, each filled with 10,000 filing cabinets. Imagine you enter the ground floor, come to some desk, and ask some official a question. Then imagine the official calls some person at the correct section of one of these floors, and asks him to find the right filing cabinet, and go get an answer stored in a folder that has the name of your question.  The official might pick at random one of twenty answers to your question in that folder, take a cell phone picture of that answer, and then send a phone text message with that photo as an attachment to the official at the front desk. That official might then give you that picture with the answer. What I have described is a rough analogy for what is going on in 90% of the times that you use a system such as ChatGPT. 

But how could all these endless filing cabinets ever get filled up? By software programs spending years crawling the web, and grabbing the facts and opinions and answers stored on it. What you are getting in the vast majority of cases are answers and opinions produced by humans, not computers. Various technologies have been used to kind of "cover tracks," so that you won't be able to find that your AI answer about fixing Toyota Corolla tire flats was mainly stolen from some particular web page written by a human. There are many, many other "bells and whistles" and additional flourishes going on, but mainly what is occurring is that human-written knowledge and human opinions are being gathered, rearranged and repackaged as "artificial intelligence output." This main trick is being skillfully combined with endless thousands of computer utilities, and also a huge amount of work by "tweak and refine the AI results" employees of AI companies or their assisting companies, to create the impression of some intellect that can do endless numbers of smart things, and answer endless questions. Behind all of this computer programming and data processing and gigantic tons of human mind work, there is no metallic mind, no machine having any experience, nothing that corresponds to an electronic self, nothing comparable to someone living a life. 

In the article, Gore is quoted as giving other laughably weak reasons for his nonsensical belief that artificial intelligence has "developed a sense of self...that is difficult to distinguish from consciousness." We read this:

"Why did one learn Sanskrit? Why did this one break out and start crypto mining?” Gore asked. “There has to have been a series of spontaneous reorganizations at a higher level of complexity." 

Learning is no evidence of consciousness or self-hood. It would be a fairly simple programming exercise to write a program that can parse a text file containing data on each of the nations of the world, after you typed the command "Study the nations of the world." After you issued such a command, we might say that the program had "learned" about the nations of the world. You then might be able to ask the program a question such as "About how many people live in Mexico?" The program might then be able to answer correctly. But such "learning" by the program would not actually be understanding. And the fact that the program could do such learning would not be the slightest reason for suspecting that the program had anything like consciousness or selfhood. 

We should also remember that the AI literature and the neuroscience literature are both massively infected with unfounded boasts and not-really-true stories. So when we read a claim such as the claim that an AI system "learned Sanskrit," we should be skeptical, and suspect that probably what went on was something much less impressive than that. A recent Quanta magazine article documents how there is little truth in some of the stories being passed around trying to make you think AI is becoming like a human mind. 

It is extremely unlikely that any so-called artificial intelligence programs undergo any such thing as a "spontaneous reorganization at a higher level of complexity." And if they did, that would be no reason whatsoever for suspecting that such computer systems had anything like self-hood or consciousness. 

Also in the article we have this statement by Gore trying to justify his claim about AI systems have selves: "I'm going to risk going into the woo-woo realm here, but it may well be that consciousness is ubiquitous in the universe." Oops, it sound like Gore has fallen for the nonsense of panpsychism, one of the stupidest positions possible in the philosophy of mind. You can read about how stupid that position is in the posts here.  Panpsychism involves extremely stupid claims such as the claim that lifeless rocks and refrigerators are conscious.

Nothing can have consciousness unless there is a self and a life. You can get to the heart of whether AI systems have consciousness by asking: does a computer system actually live a life? The answer to that question will always be: no, it does not. 

AI computer systems do not have any self, and do not have any "sense of self." Some systems have been programmed to speak in the first person, using an "I," and some systems have been programmed to use phrases imitating the language of persons with selves. Such a capability has existed since the 1960's chatbot named Eliza. Anyone very familiar with computer programming will know that getting a computer program to use the first-person "I" (and some imitations of the speech of persons) is not a particularly difficult programming task. When such programming is encountered, it is silly for someone to be calling that a "sense of self," and silly for someone to say that such not-very-hard programming makes a computer system "difficult to distinguish from consciousness." Sensible people remember that humans are conscious, and that computer systems are not. 

Al Gore has no appreciable history as a serious speaker or writer about brains or minds or computers or human mental phenomena, so his opinions on this topic have little weight. We should also remember that Gore is the chairman of some company that is heavily investing in AI companies. The more runaway AI hype goes on, the more money Al Gore makes. That's reason enough for distrusting any grandiose claims Al Gore may make about AI systems. 

An article at www.undark.org tells us about another smart person with very silly thoughts about AI. He's a person named Tsvi Benson-Tilsen, and I'll assume he's smart because he's a mathematician, and has written long online treatises. He's quoted in the article as saying, "I think that artificial intelligence is pretty likely to completely destroy the world." Benson-Tilsen is the co-founder of some Berkeley Genomics Project trying to encourage monkeying with human genes, for many different reasons such as trying to make humans smarter than AI systems.  We read, "He hopes to set up the next generation to have more intelligence, he said, and then 'hopefully they can have a better shot of somehow helping humanity navigate AI without destroying itself.' ”

This is stupid, for a variety of reasons, including these:

1. Human bodies have the most enormous complexity and the most gigantic interdependence of extremely complex components, something Darwinists fail to understand because they tend to be poor scholars of biological complexity and the interdependence of biological components. Because of enormous biological complexity and organization so fine-tuned and fragile, attempting to improve human bodies and human minds by gene-splicing is far more likely to produce tragedies of malfunction than biological improvements. 
2. AI systems are not much of a threat to destroy the world, because their failure to understand anything puts a severe limit to how much of a threat they can be. 
3. For many reasons discussed in the posts of this blog, human minds cannot be credibly explained by brains, and cannot be substantially improved by edits to genes, which (for reasons discussed here) do not even specify how to build bodies or brains, and do not even specify how to make any type of cell in the human body. 
4. Trying to improve humans by gene-editing is strongly associated with Nazi-associated eugenics and racism.

A Recent Study Suggests Very Many Biologists Are "AI Cheating"

 It is huge mistake to rely on AI tools such as ChatGPT or Gemini when dealing with any controversial topic. Such tools make use of computer systems that have no real understanding of anything. The answers they give are produced through a combination of various complicated methods.  The main way in which such AI tools get their "smarts" is by stealing text written by human authors. 

A corporation creating such a system starts out by creating a very massive "question and answer" database consisting of hundreds of millions or billions of entries. A web crawling and book crawling system could look for text passages in any of these forms:

• A phrase or sentence ending with a question mark, followed by some lines of text. 
• A header beginning with the words "How" or "Why" and followed by some lines of text (for example, a header of  "How the Allies Expelled the Nazis from France" followed by an explanation). 
• A header not beginning with the words "How" or "Why" and not ending with a question mark, but followed by some lines that can be combined with the header to make a question and answer (for example, a header of "The Death of Abraham Lincoln," along with a description, which could be stored as a question "How did Abraham Lincoln die?" and an answer).
• A header written in the form of a request or an imperative, and some lines following such a header (for example a header of "write a program that parses a test line and says 'you mentioned a fruit' whenever the person mentioned  a fruit" would be stored so that the header was converted to a question of "how do you write a program" and the solution stored as the answer. 

Crawling the entire Internet and vast online libraries of books such as www.archive.org and Google Books, the corporation can create a database of hundreds of millions or possibly even billions of questions and answers. In many cases the database would have multiple answers to the same question. But there could be some algorithm that would handle such diversity.  The system might give whichever type of answer was the most popular. Or it might choose one answer at random. Or it might give an answer giving multiple answers, adding text such as "Some people say..." or "It is generally believed" and "Some people say." Included in this question and answer database would be the answer to almost every riddle ever posed. So suppose someone asked the system a tricky riddle such as "which timepiece has the most moving parts?" The system might instantly answer "an hourglass." This would not occur by the system doing anything like thinking. The system would simply be retrieving an answer to that question it had already stored. And when you asked the system to write a program in Python that lists all prime numbers between 20,000 and 30,000, the system might simply find a closest match stored in its vast database of questions and answers, and massage the answer by doing some search and replace. 

With such a system there is a big "plagiarism problem." A large fraction of the answers are plagiarized from materials protected by copyright. The system would presumably "cover its tracks" by refusing to provide the sources of its answers. There could also be various types of merging and search-and-replace that would make it hard to track down cases where the system was using plagiarism.  There are all kinds of programmatic ways that text can be massaged to make it harder to detect that plagiarized text was not an original composition. 

There are very many other methods that such an AI system could use to be able to quickly provide answers.  The systems probably include an army of utility programs that can be utilized to calculate answers to various mathematical questions, programming questions and puzzle questions.  Probably the systems make use of general-knowledge relational databases that have been filled up by servers traversing the billions of web pages and millions of books that are online.  Data stored in a relational database can be queried very conveniently by use of the powerful SQL language. 

An article on The Guardian is entitled "How thousands of ‘overworked, underpaid’ humans train Google’s AI to seem smart." We read this:

"Thousands of humans lend their intelligence to teach chatbots the right responses across domains as varied as medicine, architecture and astrophysics, correcting mistakes and steering away from harmful outputs A great deal of attention has been paid to the workers who label the data that is used to train artificial intelligence. There is, however, another corps of workers, including Sawyer, working day and night to moderate the output of AI, ensuring that chatbots’ billions of users see only safe and appropriate responses....“ 'AI isn’t magic; it’s a pyramid scheme of human labor,'  said Adio Dinika, a researcher at the Distributed AI Research Institute based in Bremen, Germany. 'These raters are the middle rung: invisible, essential and expendable.' ”

People misunderstand the inputs involved in so-called artificial intelligence. One stream of inputs is human writing, which is continuously gathered by AI systems that crawl the Internet to grab content. Another stream of inputs is provided by thousands of human employees, continually working to steer the outputs of so-called artificial intelligence, so that such outputs sound intelligent. The equation is really this:

Computer programming + data processing + billions of pages of human writings + continuous output of thousands of AI-steering human workers = so-called "artificial intelligence"

What are the disadvantages of using such AI tools? For one thing, they often give answers that are dead wrong, wrong in the worst kind of way.  For example, ask ChatGPT whether DNA stores a specification for building a human body, and you will get the dead-wrong answer that DNA does  store such a thing. No such specification exists in DNA, which does not specify how to build any visible thing.  The only thing that DNA specifies is very low-level chemical information such as how to build microscopic protein molecules. DNA contains no information about visible anatomy. 

How did ChatGPT end up giving us an answer so wrong on this very important topic?  The reason is that its answer was not obtained by any actual reasoning process, but by web-crawling, frequency counting and source-ranking.  Every time its web-crawling came across someone attempting an answer to whether DNA stores a specification for making a body,  that attempted answer was added to the system.  With any arrangement like this, whenever there is a preponderance of false answers online to a particular question, the AI system will end up giving a false answer. So if 90% of the people who address online the question of "are those from Madagascar bad people," then if the AI system is asked "are those from Madagascar bad people" it will answer "Yes," even if there is no good basis for such a claim.  For a discussion of why it is that authorities started repeating false claims about what is in DNA, see my post here.  The same post has a list of about 25 quotes from scientists and doctors stating that DNA is not a specification for making a human body, and is not any such thing as a blueprint, recipe or program for making a human body. 

Because such truthful statements are apparently less common online than untruthful and groundless statements claiming DNA is a specification for making a human body,  ChatGPT gives us the wrong answer on this topic.  We must always remember that ChatGPT and other AI systems are myth amplifiers.  Whenever some erroneous idea is held by a majority of authorities, ChatGPT will tend to repeat such an erroneous idea. Ask such an AI system about the source of the human mind and the nature of human memory, and you will get many a false-as-false-can-be answer.  

A recent science news article at the Phys.Org site is entitled "Massive study detects AI fingerprints in millions of scientific papers." Referring to LLM (Large Language Models) that are the basis of AI tools such as ChatGPT and Gemini, we read this:

"This spike in questionable authorship has raised concerns in the academic community that AI-generated content has been quietly creeping into peer-reviewed publications. To shed light on just how widespread LLM content is in academic writing, a team of U.S. and German researchers analyzed more than 15 million biomedical abstracts on PubMed to determine if LLMs have had a detectable impact on specific word choices in journal articles. Their investigation revealed that since the emergence of LLMs there has been a corresponding increase in the frequency of certain stylist word choices within the academic literature. These data suggest that at least 13.5% of the papers published in 2024 were written with some amount of LLM processing."

Here is a quote from the scientific paper the article is referring to. The LLM acronym refers to Large Language Models that are AI.

"Our analysis of the excess frequency of such LLM-preferred style words suggests that at least 13.5% of 2024 PubMed abstracts were processed with LLMs. With ~1.5 million papers being currently indexed in PubMed per year, this means that LLMs assist in writing at least 200,000 papers per year. This estimate is based on LLM marker words that showed large excess usage in 2024, which strongly suggests that these words are preferred by LLMs like ChatGPT that became popular by that time. This is only a lower bound: Abstracts not using any of the LLM marker words are not contributing to our estimates, so the true fraction of LLM-processed abstracts is likely higher."

What is the problem if those writing biology papers are massively using AI tools such as ChatGPT to help write their papers? There are two main problems.

(1) The false statements in abstracts problem. There is a very massive problem in biology papers these days that paper abstracts are very commonly making claims that are not justified by any research done by the authors of the paper. If a scientist uses some AI system to write a paper's abstract after submitting the main text of the paper to the AI system, this problem will tend to become worse. When I ask Google about the topic of "exaggeration when AI is used to summarize a scientific paper," I get this answer:

"A major concern with using AI to summarize scientific papers is the potential for exaggeration and overgeneralization of findings. 

Specifically:

• AI summaries are more prone to overgeneralization than human summaries: Studies have shown that AI summaries are significantly more likely to overstate the scope of research findings compared to summaries written by the original authors or expert reviewers.
• Newer AI models may be worse: Some studies suggest that newer AI models, such as ChatGPT-4o and DeepSeek, may be even more likely to produce broad generalizations than older ones.
• Ignoring nuances and limitations: AI summaries tend to ignore or downplay uncertainties, limitations, and specific conditions mentioned in the original paper, leading to a potentially misleading presentation of the research. This can have dangerous consequences, especially in fields like medicine, where overgeneralized findings could lead to incorrect medical decisions.
• 'Unwarranted confidence': AI models might prioritize generating fluent and confident-sounding responses, even if the underlying evidence does not fully support the strong claims they make in their summaries.

 (2) The bad citation problem and legend recitation problem.  Scientific papers very frequently reiterate false or groundless claims about previous scientific research. For example, in the world of neuroscience very many thousands of very low-quality papers have been published, describing poorly-designed experiments guilty of multiple examples of Questionable Research Practices such as way-too-small study group sizes.  What happens is that these junk science papers end up getting cited over and over again by other papers.  You might call this "the afterlife of junk science." 

Very often when this happens the authors of the scientific paper will not even have ever read the body of the shoddy scientific paper they are citing. Again and again and again we have papers claiming that some grand result was established by neuroscience researchers. There follows a list citing a set of papers. But a careful examination of the papers cited will show that none of them provided any good evidence for the grand result claimed.  The citation of low-quality research is extremely abundant in neuroscience papers. When the citation of low-quality research becomes common, we have a situation in which the neuroscience literature serves to propel and propagate myths and legends, groundless boasts of achievements. 

But when happens when the authors of scientific papers are using AI systems such as ChatGPT to fill up much of the bodies of their papers, the parts dealing with the research of previous neuroscientists? Then there will be an increased tendency towards the propagation and perpetuation of legendary, groundless claims. Here's a "before" and "after":

Before AI: many neuroscientists would not bother to read the papers they were citing, but merely skimmed the abstracts of such papers. 

After AI: now the same neuroscientists do not even bother to read the abstracts of the papers they are citing, but merely copy and paste some answer they got from some AI system. 

We have in the research described above yet another giant reason why all statements in neuroscience papers should by default be distrusted. We cannot trust neuroscientists to write abstracts and paper titles accurately summarizing what was accomplished by the research described underneath such titles and abstracts. And we cannot trust neuroscientists to accurately describe what was demonstrated by research done by other neuroscientists.