Alarming: Fraud spreads in Science -- and I fear it will become worse

Science has a big problem and it’s been getting rapidly worse in the past two years or so, to no small part because of recent advances in artificial intelligence. Fraudulent papers are getting published more than ever, and the fraudsters are getting increasingly aggressive. In this episode I want to give you an update on the recent developments.

🔗 Join this channel to get access to perks ➜

00:00 Intro
01:42 Fraud in Science

#science #sciencenews

[soaringeagle]

What may be even worse is when the president of the university won't let you publish a good paper because it refutes what the donor is trying to prove and the president doesn't want the university to lose the research grant.

[patrickdaly]

I think this is a great example of Goodhart's law. "When a measure becomes a target, it ceases to be a good measure." If number of papers is the measure, then science isn't even the goal anymore.

[jannieschluter]

I am PhD Chemist. what you are waking up to is the reason why I am unemployed and proud of it. There is no good reason to be a scientist in this corrupt scientific society.

[Name-otxw]

My main problem with scientific publishing is that there is very little focus on replication. I'm taking bets on how many studies we could throw away if we were to make a serious effort towards replication, I put myself down for about half.

[glennnelson-rt]

I worked in the physics dept when an undergraduate, so I knew many of the grad students and professors.
It was common knowledge that one prolific professor who was a phys rev referee would reject a submitted paper and then
assume the research himself, another egregious type of fraud.

[kdbublitz]

I don't think people have ever lost this much faith in our institutions.. It is absolutely frightening.

[jackh]

Thanks for this Sabine. I decided against pursuing a career in academia when I interned with an academic at the University of Oxford last summer who admitted to flagrantly falsifying validation stats for his papers as he "couldn't be bothered" to derive them honestly but was "pretty sure" his models were going to be 90+% accurate anyhow. I was visibly aghast at this - he clearly sensed this and spent the remainder of my placement chastising me for being "arrogant" and "acting like I know better than everyone" and then didn't respond to my reference request. Seems bullies are rewarded in academic power structures even more so than in corporate ones. I'm now very cynical about novel findings in papers and wait for them to be widely replicated before I believe them.

[lwmarti]

I used to review papers for a few publications. Once I noted how a particular reference didn't actually say what the referring author said it did. That particular topic was never even mentioned in the reference. After that, that journal stopped asking me to review papers. And, yes, the paper with the suspicious reference got published anyway. 🤔

[appalachianamerican]

It's been bad for decades. Now it's so obvious it can no longer be hidden.

[deltalima]

30 years ago this was already a problem. I am happy to see it finally getting some publicity. Thanks Sabine!
👍

[FourthRoot]

We shouldn't talk about science as though it's a community, an industry, or a collective of any kind. It's a method. What you're referring to is academia, which is not always scientific, even in their "sciences" departments.

[buybuydandavis]

Pournelle's Iron Law of Bureaucracy:
In any bureaucracy, the people devoted to the benefit of the bureaucracy itself always get in control and those dedicated to the goals the bureaucracy is supposed to accomplish have less and less influence, and sometimes are eliminated entirely.

[macbitz]

Seems there is no area of human endeavour where crime doesn't play a significant part.

[shadowdragon]

The fact that there is a monetary incentive to publish as many papers as possible is a problem in and of itself. Ideally there should be incentives for quality over quantity as well as attempting to reproduce the results of existing studies.

[elfpimp]

I guarantee you that for every fraud found more than a dozen gets through. This is as old as humanity.

[faolitaruna]

It should be said that last year more than 5 million scientific papers were published. This is a problem in itself.

[kennethnichols]

Confirmation of the old axion “ Show me the incentives and I’ll show you the outcome”.

[kurtisb]

Meanwhile…. Trust the science bro.

Long ago I came to realize the wisdom in Charlie mungers quote “you show me the incentive and I will show you the outcome”. Anyone who thinks otherwise is a fool.

[manhuawang]

Luckily this doesn't apply to fields related to public policy with massive budgets and far-reaching global consequences.

[JohnJaneson]

I have to read scientific journals regularly for my work in military RnD. I notice that newer papers are full of regurgitations and fluff, no new knowledge at all.

They can pass their thesis defenses, but when read for real industrial usage, the lies become obvious.