Danny Maupin
@dmaupin.bsky.social
🔬Research Fellow in Health Science University of Surrey
🩺Specialist Vestibular Physiotherapist
🩺Specialist Vestibular Physiotherapist
I read somewhere on here an idea that publishing scientists should be registered like health professionals. Its an interesting idea and wouldn't stop all bad behaviour but if it's a) independent, international organisation, and b) can pull registrations for significant abuses I think its interesting
November 6, 2025 at 1:30 PM
I read somewhere on here an idea that publishing scientists should be registered like health professionals. Its an interesting idea and wouldn't stop all bad behaviour but if it's a) independent, international organisation, and b) can pull registrations for significant abuses I think its interesting
Just another way AI is overwhelming academia - see our preprint here about a 17x increase over the same period in redundant publications using NHANES
www.medrxiv.org/content/10.1...
www.medrxiv.org/content/10.1...
Dramatic increases in redundant publications in the Generative AI era
Background Redundant publication, the practice of submitting the same or substantially overlapping manuscripts multiple times, distorts the scientific record and wastes resources. Since 2022, publicat...
www.medrxiv.org
November 5, 2025 at 5:17 PM
Just another way AI is overwhelming academia - see our preprint here about a 17x increase over the same period in redundant publications using NHANES
www.medrxiv.org/content/10.1...
www.medrxiv.org/content/10.1...
Sound nice, but usually the technology used by bad actors outpaces that used by the regulators so it will need to be implemented effectively
November 5, 2025 at 2:47 PM
Sound nice, but usually the technology used by bad actors outpaces that used by the regulators so it will need to be implemented effectively
Will be influenced by a crap paper?
I do think it is important because we need a system that fosters trust, particularly from those outside academia, without a) blanket rejection and b) putting more workload on individuals to self police paper mills and other bad actors.
The use of AI tools
I do think it is important because we need a system that fosters trust, particularly from those outside academia, without a) blanket rejection and b) putting more workload on individuals to self police paper mills and other bad actors.
The use of AI tools
November 5, 2025 at 2:47 PM
Will be influenced by a crap paper?
I do think it is important because we need a system that fosters trust, particularly from those outside academia, without a) blanket rejection and b) putting more workload on individuals to self police paper mills and other bad actors.
The use of AI tools
I do think it is important because we need a system that fosters trust, particularly from those outside academia, without a) blanket rejection and b) putting more workload on individuals to self police paper mills and other bad actors.
The use of AI tools
I think the issue isn't with good papers or authors using the tool effectively. The issue would be more the onslaught of slop that is published. Peer review was already overwhelmed before this. If you rely on finding it post "publication" how long will that take? How many news, media, reveiws
November 5, 2025 at 2:47 PM
I think the issue isn't with good papers or authors using the tool effectively. The issue would be more the onslaught of slop that is published. Peer review was already overwhelmed before this. If you rely on finding it post "publication" how long will that take? How many news, media, reveiws
I agree blanket rejection is dumb.
I don't think using AI is the answer either though. I believe the bad actors and their AI would be able to stay ahead of the good AI so a lot would still go through. Do you know of any findings showing AI effectively putting a cap on the AI slop?
I don't think using AI is the answer either though. I believe the bad actors and their AI would be able to stay ahead of the good AI so a lot would still go through. Do you know of any findings showing AI effectively putting a cap on the AI slop?
November 5, 2025 at 2:20 PM
I agree blanket rejection is dumb.
I don't think using AI is the answer either though. I believe the bad actors and their AI would be able to stay ahead of the good AI so a lot would still go through. Do you know of any findings showing AI effectively putting a cap on the AI slop?
I don't think using AI is the answer either though. I believe the bad actors and their AI would be able to stay ahead of the good AI so a lot would still go through. Do you know of any findings showing AI effectively putting a cap on the AI slop?
Discovering Statistics by Andy Field - he does a great job introducing beginner stats concepts in a way that's not too heavy. Great entry for those with non-stats/math backgrounds
October 20, 2025 at 6:09 AM
Discovering Statistics by Andy Field - he does a great job introducing beginner stats concepts in a way that's not too heavy. Great entry for those with non-stats/math backgrounds
Employment benefits to underemployed workers to benefit their mental health. Now time to find a home for this piece!
#academicsky #opensci #episky
#academicsky #opensci #episky
October 18, 2025 at 9:06 AM
Employment benefits to underemployed workers to benefit their mental health. Now time to find a home for this piece!
#academicsky #opensci #episky
#academicsky #opensci #episky
Impactful publications but that is hard to measure
October 12, 2025 at 9:09 AM
Impactful publications but that is hard to measure
Particularly if all that is being churned out is junk or fast churn science. It's easy to use AI to develop about x predicts y and another paper about z predicts why or redundant publications. This has been a problem before AI but is accelerated now. Looking forward to it publishing more robust /
October 12, 2025 at 9:09 AM
Particularly if all that is being churned out is junk or fast churn science. It's easy to use AI to develop about x predicts y and another paper about z predicts why or redundant publications. This has been a problem before AI but is accelerated now. Looking forward to it publishing more robust /
No I don't think it would be, just write things vague enough
October 10, 2025 at 6:57 PM
No I don't think it would be, just write things vague enough
You think he will bother pre-registering? I'm doubtful
October 10, 2025 at 6:11 PM
You think he will bother pre-registering? I'm doubtful
I may be confusing the proof-evidence distinction so please educate me if I am, but I don't think we make proof
October 10, 2025 at 3:59 PM
I may be confusing the proof-evidence distinction so please educate me if I am, but I don't think we make proof
The hypothesis remark was specific to a reply about looking for a hypothesized effect.
I don't think you make proof if you're doing research without a hypothesis though? You are discovering proof that exists. I think people take issue with the word make particularly with someone openly biased
I don't think you make proof if you're doing research without a hypothesis though? You are discovering proof that exists. I think people take issue with the word make particularly with someone openly biased
October 10, 2025 at 3:56 PM
The hypothesis remark was specific to a reply about looking for a hypothesized effect.
I don't think you make proof if you're doing research without a hypothesis though? You are discovering proof that exists. I think people take issue with the word make particularly with someone openly biased
I don't think you make proof if you're doing research without a hypothesis though? You are discovering proof that exists. I think people take issue with the word make particularly with someone openly biased
The same page about the idea being ridiculous
October 10, 2025 at 3:28 PM
The same page about the idea being ridiculous
So I still take issue with the make proof aspect, we could test the hypothesis, collect data all that but make proof has a different connotation especially when said by someone openly biased.
I think this conversation is in good fun as it's interesting debating science philosophy and glad we are /
I think this conversation is in good fun as it's interesting debating science philosophy and glad we are /
October 10, 2025 at 3:28 PM
So I still take issue with the make proof aspect, we could test the hypothesis, collect data all that but make proof has a different connotation especially when said by someone openly biased.
I think this conversation is in good fun as it's interesting debating science philosophy and glad we are /
I think this conversation is in good fun as it's interesting debating science philosophy and glad we are /
But we shouldn't look for hypothesized effects. We make a hypothesis and then we test it. That's different. And even the species example the more I think about it isn't making proof, we are looking for evidence yes but we don't make the proof. The proof has existed previously we just didn't see it /
October 10, 2025 at 3:28 PM
But we shouldn't look for hypothesized effects. We make a hypothesis and then we test it. That's different. And even the species example the more I think about it isn't making proof, we are looking for evidence yes but we don't make the proof. The proof has existed previously we just didn't see it /
using science as a monolith when there are examples that prove otherwise so they probably should've been more specific
October 10, 2025 at 2:23 PM
using science as a monolith when there are examples that prove otherwise so they probably should've been more specific
I'd say that's a little disingenuous because we aren't looking at discovering new species, this is specifically health research, and not health research that is looking for a new disease, but looking at causality of a condition with multiple confounders. The other poster is being too broad /
October 10, 2025 at 2:23 PM
I'd say that's a little disingenuous because we aren't looking at discovering new species, this is specifically health research, and not health research that is looking for a new disease, but looking at causality of a condition with multiple confounders. The other poster is being too broad /
Not that it always is as researchers often have biases, but this feels like poor wording from individuals that don't seem concerned with gold standard science despite them saying they are
October 10, 2025 at 12:47 PM
Not that it always is as researchers often have biases, but this feels like poor wording from individuals that don't seem concerned with gold standard science despite them saying they are
I think people take issue with the wording due to the belief that they are going to interfere with any data or conclusions to back up their point, thus making proof.
If I understand correctly you are saying science is always looking to make proof, but ideally this will be proof for or against /
If I understand correctly you are saying science is always looking to make proof, but ideally this will be proof for or against /
October 10, 2025 at 12:47 PM
I think people take issue with the wording due to the belief that they are going to interfere with any data or conclusions to back up their point, thus making proof.
If I understand correctly you are saying science is always looking to make proof, but ideally this will be proof for or against /
If I understand correctly you are saying science is always looking to make proof, but ideally this will be proof for or against /