Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
AI Science

Scientific Journal Publishes AI-Generated Rat With Gigantic Penis (vice.com) 72

Posted by BeauHD from the worrying-incidents dept.
Jordan Pearson reports via Motherboard: A peer-reviewed science journal published a paper this week filled with nonsensical AI-generated images, which featured garbled text and a wildly incorrect diagram of a rat penis. The episode is the latest example of how generative AI is making its way into academia with concerning effects. The paper, titled "Cellular functions of spermatogonial stem cells in relation to JAK/STAT signaling pathway" was published on Wednesday in the open access Frontiers in Cell Development and Biology journal by researchers from Hong Hui Hospital and Jiaotong University in China. The paper itself is unlikely to be interesting to most people without a specific interest in the stem cells of small mammals, but the figures published with the article are another story entirely. [...]

It's unclear how this all got through the editing, peer review, and publishing process. Motherboard contacted the paper's U.S.-based reviewer, Jingbo Dai of Northwestern University, who said that it was not his responsibility to vet the obviously incorrect images. (The second reviewer is based in India.) "As a biomedical researcher, I only review the paper based on its scientific aspects. For the AI-generated figures, since the author cited Midjourney, it's the publisher's responsibility to make the decision," Dai said. "You should contact Frontiers about their policy of AI-generated figures." Frontier's policies for authors state that generative AI is allowed, but that it must be disclosed -- which the paper's authors did -- and the outputs must be checked for factual accuracy. "Specifically, the author is responsible for checking the factual accuracy of any content created by the generative AI technology," Frontier's policy states. "This includes, but is not limited to, any quotes, citations or references. Figures produced by or edited using a generative AI technology must be checked to ensure they accurately reflect the data presented in the manuscript."

On Thursday afternoon, after the article and its AI-generated figures circulated social media, Frontiers appended a notice to the paper saying that it had corrected the article and that a new version would appear later. It did not specify what exactly was corrected. UPDATE: Frontiers retracted the article and issued the following statement: "Following publication, concerns were raised regarding the nature of its AI-generated figures. The article does not meet the standards of editorial and scientific rigor for Frontiers in Cell and Development Biology; therefore, the article has been retracted. This retraction was approved by the Chief Executive Editor of Frontiers. Frontiers would like to thank the concerned readers who contacted us regarding the published article."
This discussion has been archived. No new comments can be posted.

Scientific Journal Publishes AI-Generated Rat With Gigantic Penis More

Scientific Journal Publishes AI-Generated Rat With Gigantic Penis

Comments Filter:

  • Appears to be a really low-quality, no impact factor online publication. What's the big deal again?

    • ...What's the big deal again?

      Perhaps that the member is twice the size of the male's body? Have another look at the AI generated image and if you still have questions, ask your parents.

      • Re: (Score:2)

        by Calydor ( 739835 )

        If it was there as a magnification with the rest of the body simply to confirm the species we're talking about that'd be one thing, but it doesnt' seem to be the case here.

      • Re: (Score:2)

        by piojo ( 995934 )

        I didn't think it was that bad. It is an enlarged cutaway diagram. The fact that it's a penis just makes it mildly amusing.

        • Re: (Score:3)

          by XXongo ( 3986865 )

          I didn't think it was that bad. It is an enlarged cutaway diagram.

          It was a nonsense diagram with parts labeled with nonsense words like “iollotte sserotgomar cell” and “testtomcels."

          it bears no relationship to rat anatomy, or for that matter to the reproductive anatomy of any mammal.

          • Re: (Score:2)

            by piojo ( 995934 )

            I see. My point, based on who I replied to, was that this style of enlarged cutaway diagram is appropriate, but I did not analyze it for accuracy of content.

      • Re: (Score:2)

        by dbialac ( 320955 )

        ...What's the big deal again?

        Perhaps that the member is twice the size of the male's body? Have another look at the AI generated image and if you still have questions, ask your parents.

        I think you're taking the possibility of a pun too seriously.

    • The big deal is that it is going to get bigger and better. People are already stupidly believing every piece of trashy vaccine misinformation aunt Sharon is putting out on facebook. The last thing we need is every second novel or scientific journal being written by Bot Lieberman

    • welcome our penis-joke loving AI overlords

      No telling what genetic modifications they have planned for us humans

    • Yes, calling anything published by Frontiers a scientific journal is rather stretching the term.

  • it's because they outsourced the job to India!

  • Given the niche nature of the paper i could understand if regular editors missed things like this, but the whole point of "peer review" is that other people working in the field who can understand the content should be checking this. Clearly that hasn't happened.

    So if something claims to be peer reviewed, how do you know if it really has? Unless there's an obvious mistake like this one, you have no idea.

    • There are lots of scam science journals. Basically outlets like this allow you to publish anything for a fee. This helps boost a researchers publishing rate which helps them with grant money. The whole enterprise of science has become like this due to government funding putting priorities on quantities.

      • Peer review and editing are also thankless jobs. It can take a while to process a research paper, and it is not like the workload ever gets easier except on holidays maybe. At some point people just stop caring and treat the scientific field as any other 9-5 job.

        • Re: (Score:3)

          by gtall ( 79522 )

          Yes, it does take a long time to process a research paper. I was once asked to review a paper on ring theory for a logic journal. I'm no ring theorist but I do know algebra. So I cracked open a ring theory book and got to work. Once I had decoded the paper and was able to recreate the missing proofs (proofs are sometimes left out of papers), I found the author had made a simple assertion early on. I could not recreate the proof. So I asked for the proof. W/hat I got back was a proof by invocation of a highe

          • Re: Peer review? (Score:4, Informative)

            by burtosis ( 1124179 ) on Friday February 16, 2024 @08:35AM (#64244342)
            Peer review is also often handed down to graduate students who are unpaid for the work. Salaried at 20 hours or even less is the norm and they are expected to do 40 hours work not including peer review which is just thrown on top of a monstrus pile. Back when I had this thankless job the running joke was we were paid significantly less than McDonald’s employees because on an hourly basis it was true. For the crap thrown atop the pile of unpaid work we would quip about doubling pay for people that did a good job or halving it for those who didn’t. Sadly this meant that peer review was often just a cursory reading and a few points raised because there was literally no time for anything else before the next shovel full of crap hit us in the face.

          • Re: (Score:2)

            by guruevi ( 827432 )

            Most peer review these days is done on a budget, in the field I work you get an honorarium of about $700, so peer review is expensive on the side of the researcher, who now has to shell out $2100 for just 3 reviewers, then who wants to do 3 months worth of work for $700?

            Peer review is not a replication study (which is what you did). Peer review is a few hours of reading the paper and making sure there are no glaring issues, you may ask methodology question but it would be rare for someone to actually critic

      • A few years ago I was looking into Single Event Upset in certain devices that were deployed in terrestrial applications. Noting that failures appeared to be more frequent at higher-altitude sites, informed by a certain homework problem in third semester undergrad physics that introduced relativistic time dilation that allows muons from upper atmosphere cosmic particle interactions to be observed at Earth's surface, and after doing a little reading on SEU, I searched for papers describing experimental resul

  • ..., who said that it was not his responsibility to vet the obviously incorrect images. ... "As a biomedical researcher, I only review the paper based on its scientific aspects"

    An included image is part of the paper. Does he also not bother checking graphs for accurate portrayal of data?

    I suppose there's a pro tip there. Add any image you like. As long as you say Midjourney made it, it will pass his review. Test this theory with a generated pic involving his mother getting airlocked.

    • Re: (Score:3)

      by Calydor ( 739835 )

      Why generated?

      • I suppose there's a pro tip there. Add any image you like. As long as you say Midjourney made it, it will pass his review.

        Why generated?

        I’m assuming a generated image was way easier to make than getting the camera and body position just right much less actually the difficulty of fabricating a giant silicone rat phallus and gluing it on.

    • Re: (Score:2)

      by dbialac ( 320955 )
      In this case I think the error boiled down to the vast information that the AI tool got was heavily weighted towards information about online material on human penises.

      • Re: (Score:2)

        by XXongo ( 3986865 )

        In this case I think the error boiled down to the vast information that the AI tool got was heavily weighted towards information about online material on human penises.

        It's not accurate about the interior structure of human penises either. Other than the fact they they are cylindrical, every single detail is wrong (and the labels are hilarious).

    • Re: (Score:2)

      by ceoyoyo ( 59147 )

      It seems pretty unlikely he reviewed it. Reviewers love to be super critical. He probably passed it off to a grad student or even an undergrad who did a half-assed job because they a) weren't getting paid for it and b) weren't even asked in the first place.

  • Pics or didn't... (Score:4, Funny)

    by Opportunist ( 166417 ) on Friday February 16, 2024 @05:48AM (#64244090)

    uh...

    Never mind!

  • Recently there was a news article about how there has been a huge uptick of fraudulent academic papers in China because there is such intense pressure there to put them out so they can advance their careers. Could this be yet another one of these fraudulent papers?
  • this is the result... Utter gibberish

  • Wow, just wow. While the figure does appear to be somewhat in poor taste, it isn't supposed to be representative, but, instead to provide better detail of the subject of the paper. If the rat's kidney, or foot, or heart were enlarged like that, to show detail, do you think the image would have been subject to such attention?

    Using AI-generated graphics in this case is no different from farming out your diagrams to a human-run artistic service who knows little more than the AI would about exactly what they

    • At the end of the day, the image used was absolutely garbage. Not just the art itself but also the text being butchered to the point of being illegible.
      And I doubt people would give it a pass if this was the product of human artistry. People make fun of bad art all the time even before generative models came onto the scene. The context of it being used in a research paper is arguably the more problematic aspect.

    • I would contest and say that the artist would know what a penis is. The scientific research into the topic readily available on a certain category of frequently sought out websites. The photographic research material is also kept locally for long term storage and is thus readily available on many home computers and smartphones. I otherwise agree. The enhanced size, while initially quite comical, provides details that a to scale image with the rat simply cannot provide. It's also great because it provides fo
  • Ever wonder when you see those TV commericials that claim "researched for over 50 years", or "clinically tested" how they can make those claims? Well, this isn't one of those as those two statements tell you nothing about any research nor clinical results. lol But these articles can now be referenced adding the legitimacy that people want to see. Nearly 100% of items being hocked on TV or websites like Amazon, Tik Tok, F Book have found ways to use this fake data by simply referencing the headline. Nobody

  • From the actual (retracted) paper:
    RECEIVED 17 November 2023
    ACCEPTED 28 December 2023

    Anything from this open-access publisher is crap. They accept crap papers, and they do it in an incredible short period to make authors happy. No serious review can be made in such short period, and it is really strange to have a paper accepted in the first round of review -- unless, of course, the review process is crap. Also note they do this so fast to charge the authors with the publishing fees [frontiersin.org] (yes, authors pay to have

    • Re: (Score:2)

      by ceoyoyo ( 59147 )

      It was a marvellous triumph by the publishing industry to head off complaints about how much they charge for access AND convince everyone that a bribery based scheme was both open and more ethical.

  • "It's unclear how this all got through the editing, peer review, and publishing process."

    Is it unclear?
    Really?
    Or is it that we are uncomfortable when we are compelled to admit that there is no meaningful "process " or filter on these publications, casting doubt into the veracity and of the entire system (and meanwhile the authority of its results)?

    I submit: given the facts, which is more likely to be true?

  • maybe this will finally cause some change (Score:5, Insightful)

    by v1 ( 525388 ) on Friday February 16, 2024 @07:40AM (#64244254) Homepage Journal

    It's unclear how this all got through the editing, peer review, and publishing process.

    Perhaps AI sending them laughably absurd things that they "automatically review and publish" is what it will take to make them start actually REVIEWING things before they publish them.

    Getting your work peer reviewed nowadays no longer means anything because they'll publish ANYTHING. And AI is making them a laughing stock.

    • It's unclear how this all got through the editing, peer review, and publishing process.

      Perhaps AI sending them laughably absurd things that they "automatically review and publish" is what it will take to make them start actually REVIEWING things before they publish them.

      Getting your work peer reviewed nowadays no longer means anything because they'll publish ANYTHING. And AI is making them a laughing stock.

      At least one of the peer reviewers said he just ignored the images (it's mentioned in the summary), and focused only on the scientific content. So I don't think it's fair to say that it wasn't reviewed, just that the reviewer didn't figure illustrations were relevant. No one seems to be claiming that the meat of the paper is invalid.

  • The photos of my gigantic penis, published without my permission by a former girlfriend, were released before AI was invented.

    I felt it was important to ensure no misunderstandings arise in the wake of this unfortunate story. Thank you for your attention. That is all.

  • ... about the publisher [wikipedia.org], which says they publish "over 185" journals, and employ over 1400 people. The article is gently but firmly negative about Frontiers Media.

  • Dai Chinpo?
    Is the article AI generated too?

  • As I understand it, cheating and dishonesty on academic papers is endemic in some parts of the world, fuelled by ridiculous incentives that reward quantity over quality. Adding AI into this mix will only speed up the race to the bottom although may also provide some useful tools to vet articles
  • This should rather be called degenerative AI.
  • You cannot demand from non-interested partner to stop abusing the system. If you block it off from the system, you loose possibility to spy over it, but it keeps access to your databases and continues to take information and hints for its own research. Flooding science journals with fake or just trash publications is the best way to undermine research of the competitor and drive away its research focus and money to wrong direction. It is a regular form of industrial sabotage. Plus it gives possibility to ma

  • Well, it's been retracted, but you can still download it. I don't actually see a problem with the rat picture. The penis part is just an exploded enlarged view with cutouts to show the features they are talking about. The article might be crap research, but there is nothing absurd about the diagram.

  • ... as dominatixes everywhere applauded the availability of giant rat penises...
  • What does that even mean now?

    • Re: (Score:2)

      by Briareos ( 21163 )

      It was a simple typo; it really should have said "pecker review" - an easy mistake to make...

  • I remember, in 1981, when we had to dissect a rat as part of Biology class. One of my classmates asked where its penis was as it was obvious by its prominant testicles that it was a male. The teacher rummaged around its genital area with a scalpel for a few minutes before finding it and cutting it off. It was tiny, it looked very much like some matted fur, no wonder we couldn't find it.

  • the article got reviewed before publication, but all kind of excuses were given why it still ended up on their site.
    maybe they can use AI to help peer review the articles?

Slashdot Top Deals

You are always doing something marginal when the boss drops by your desk.

Close