Jack Ryan 🙏🧪
@jacksonwryan.com
Not a spy.
Investigative / science journalist.
President, Science Journalists Association of Australia.
I write about people, science, integrity, dodgy stuff, video games, art etc.
Investigative / science journalist.
President, Science Journalists Association of Australia.
I write about people, science, integrity, dodgy stuff, video games, art etc.
Yeah. I suspect DeepSeek or another LLM is built into the WPS or translation plugins being used -- this may only be recently updated and That's why we're seeing this from sort of August onwards. Still digging!
January 12, 2026 at 9:17 PM
Yeah. I suspect DeepSeek or another LLM is built into the WPS or translation plugins being used -- this may only be recently updated and That's why we're seeing this from sort of August onwards. Still digging!
They are definitely in the PDF versions online for this one, so I imagine they are in print, too.
January 12, 2026 at 9:45 AM
They are definitely in the PDF versions online for this one, so I imagine they are in print, too.
It looks like a potential translation tool issue. However, I have found it in places where I wouldnt expect if it was only a translation tool issue... More next week 👀
January 12, 2026 at 8:15 AM
It looks like a potential translation tool issue. However, I have found it in places where I wouldnt expect if it was only a translation tool issue... More next week 👀
It does seem impressive, though it pulled most of that thinking from my own postulating and Leon's first discoveries, I think. The idea of a firmware update is what I raise in the piece, and then it spat that back at you. But it did find it in official documents, so it's certainly found some things!
January 12, 2026 at 3:57 AM
It does seem impressive, though it pulled most of that thinking from my own postulating and Leon's first discoveries, I think. The idea of a firmware update is what I raise in the piece, and then it spat that back at you. But it did find it in official documents, so it's certainly found some things!
Of course, the bigger problem is how some of these peer review processes work in allowing nonsensical sentences. It would be interesting to know at what step the translation tool started throwing in the species names, but I suspect it was prior to revisions and acceptance!
January 12, 2026 at 3:37 AM
Of course, the bigger problem is how some of these peer review processes work in allowing nonsensical sentences. It would be interesting to know at what step the translation tool started throwing in the species names, but I suspect it was prior to revisions and acceptance!
Broussonetia, in particular, appears in the strangest locations. This paper, for instance, has the reference 11 times, as well as Parazacco: www.sciencedirect.com/science/arti...
Bibliometric analysis of mesenchymal stem cell-derived extracellular vesicles in the treatment of osteoarthritis
This study adopts bibliometric and visualization analysis to systematically investigate the research hotspots and emerging trends in the treatment of …
www.sciencedirect.com
January 12, 2026 at 3:01 AM
Broussonetia, in particular, appears in the strangest locations. This paper, for instance, has the reference 11 times, as well as Parazacco: www.sciencedirect.com/science/arti...
Yes, there's a bit more behind the scenes here: re-translations by authors I contacted and that did respond did not raise the issues. There's also a few times where "Homo sapiens" has been replaced for words like personal, and I suspect the symbol "人" is responsible for this.
January 12, 2026 at 2:46 AM
Yes, there's a bit more behind the scenes here: re-translations by authors I contacted and that did respond did not raise the issues. There's also a few times where "Homo sapiens" has been replaced for words like personal, and I suspect the symbol "人" is responsible for this.
I will add that I contacted authors and one did refer to an erroneous "AutoText" template. But the response I received, itself, seemed AI generated. Many other authors I contacted have not responded. I am leaning toward this is a tool mistranslation -- but it also exposes some poor peer review.
January 12, 2026 at 2:15 AM
I will add that I contacted authors and one did refer to an erroneous "AutoText" template. But the response I received, itself, seemed AI generated. Many other authors I contacted have not responded. I am leaning toward this is a tool mistranslation -- but it also exposes some poor peer review.
I think I totally misinterpreted what you're saying so deleted the comment (send those news articles -- are you suggesting that AI generated them based on school closure dates or something?)
January 12, 2026 at 2:13 AM
I think I totally misinterpreted what you're saying so deleted the comment (send those news articles -- are you suggesting that AI generated them based on school closure dates or something?)
Yes! Definitely one line of inquiry. Likely is caused by machine translation. But these answers don't tackle the scientific integrity / peer review issues (ie, why are they appearing in the literature and at such high frequency). If you speak Chinese, have you confirmed the outputs here are correct?
January 12, 2026 at 2:11 AM
Yes! Definitely one line of inquiry. Likely is caused by machine translation. But these answers don't tackle the scientific integrity / peer review issues (ie, why are they appearing in the literature and at such high frequency). If you speak Chinese, have you confirmed the outputs here are correct?
🧪 In this table, for instance, there's a ton of the fish names in the final row ... strange! This parazacco has appeared across dozens of papers, but there are a few more species names popping up
nobreakthroughs.substack.com/p/is-this-fi...
nobreakthroughs.substack.com/p/is-this-fi...
January 12, 2026 at 12:32 AM
🧪 In this table, for instance, there's a ton of the fish names in the final row ... strange! This parazacco has appeared across dozens of papers, but there are a few more species names popping up
nobreakthroughs.substack.com/p/is-this-fi...
nobreakthroughs.substack.com/p/is-this-fi...
It could be mistranslations, for sure. That could explain why it's present throughout web novels by this author (www.webnovel.com/profile/4503...) or on Alibaba listings, but the mistranslated part has been difficult to pin down. Chinese colleagues I contacted haven't been able to work it out yet.
January 12, 2026 at 12:26 AM
It could be mistranslations, for sure. That could explain why it's present throughout web novels by this author (www.webnovel.com/profile/4503...) or on Alibaba listings, but the mistranslated part has been difficult to pin down. Chinese colleagues I contacted haven't been able to work it out yet.
Real ones know Dazzleland and Port Power in the 2021 home prelim
January 12, 2026 at 12:02 AM
Real ones know Dazzleland and Port Power in the 2021 home prelim
The same Melissa McIntosh kicked out of QT for saying the Labor government is not doing enough to ban social media? Who would have thought!?
January 9, 2026 at 11:20 PM
The same Melissa McIntosh kicked out of QT for saying the Labor government is not doing enough to ban social media? Who would have thought!?
You're also in there now Sophie!
January 5, 2026 at 10:34 PM
You're also in there now Sophie!
Well, there was about 2 minutes where it seemed like it might be competitive ☢️
January 3, 2026 at 9:25 PM
Well, there was about 2 minutes where it seemed like it might be competitive ☢️
Can the authors comment on these similarities? 😆😆 Happy New Year!
January 1, 2026 at 10:51 AM
Can the authors comment on these similarities? 😆😆 Happy New Year!
Journalists goldmine today! What riches will we discover???
December 31, 2025 at 5:14 AM
Journalists goldmine today! What riches will we discover???
Great! That's a lot of movers and shakers! How many of the 250 were science journalists? Looking forward to a greater diversity of journalism beats next year.
December 21, 2025 at 12:03 AM
Great! That's a lot of movers and shakers! How many of the 250 were science journalists? Looking forward to a greater diversity of journalism beats next year.
Very misleading graphic. The video message is more important than the graphic but this is bad data representation. At a glance, you can't get any real information from this -- and the information you might get doesn't reflect the coverage
December 19, 2025 at 6:03 AM
Very misleading graphic. The video message is more important than the graphic but this is bad data representation. At a glance, you can't get any real information from this -- and the information you might get doesn't reflect the coverage