> It seems like a strange thing to take someone with a long and respected career and subject them to what would essentially be a Western blot and photomicrograph audit before offering them a big position.
This is absolutely something that we should routinely be doing, though.
It's pretty similar to the level of distrust in the software engineering job interview process.
Pick your poison, to some extent. Better would be to not have to do it after-the-fact, but to vet better at every intermediate step, but it's hard. Just a very difficult people problem.
Agreed. There!s uproar over coding interviews, which makes no sense to me. We give easy peasy code reviews to smoke check claimed skills. 4 of 5 candidates do absolutely terribly on very easy stuff, relative to their claimed skillset. No, our bar isn’t high—fraud (resume fraud) is sadly very real.
yeah it sounds a little bit absurd to me. It's just basic due diligence. You don't not run a background check on a potential employee just bc their resume looks good and they got a reference. In those cases you still go, "Annoying we have to wait because we want this person on board NOW and it's a fairly shallow investigation that 99% of the time doesn't reveal anything even if there is something, but it's the standard procedure."
Maybe we need instruments that sign results cryptographically and use a blockchain mechanism to establish provenance. We should have cameras that can establish that published images have not been modified (or at least provide raw and adjusted pairs--digital radiology has the concept of a "presentation state" that I think could work).
In theory at least research should be auditable to a lab notebook. The problem with photos and such is you can't tell if it was modified before it was pasted to the page and large datasets just can't be put in to paper. And electronic notebooks I've used tend to be even more annoying than paper (too rigidly formatted and not adaptive to workflow optimization but it's difficult to explain).
Anyway those sorts of things that establish provenance should also protect against deep learning. You may be able to create fake data, but can you deep fake data that was signed by Nikon device #XYZ with cryptographicly confirmed hashes published to a blockchain 3 years ago at the time the data was generated?
There are unfortunately very rarely consequences for academic fraud. It's not just that we only catch a small fraction — mostly the most brazen image manipulation — but these cases of blatant fraud happen again and again, to resounding silence.
Ever so rarely, there may be an opaque, internal investigation. Mostly, it seems that academia has a desire to not make any waves, keep up appearances, and let the problem quiet down on its own.
The people doing the investigation have a vested interest in keeping it quiet.
It's like the old quote... "If you commit fraud as an RA that's your problem. If you commit fraud as the head of department that's the university's problem."
And occasionally a grad student who discovers academic dishonesty, and complains internally (naively trusting administrators to have humility and integrity), has their career ended.
I suppose a silver lining to all the academic fraud exposés of the last few years is that more grad students and faculty now know that this is a thing, and one that many will try to cover up, so trust no one.
Another silver lining might be that fellow faculty are more likely to believe an accusation, and (if they are one of the awful people) less likely to think they can save funding/embarrassment/friend by neutralizing the witness.
(ProTip: If the success of your dishonesty-reporting approach is predicated on an internal administrator having humility and integrity, realize that those qualities are the opposite of what has advanced a lot of academic careers.)
The amazing part about this to me is that the only reason the authors were caught is image manipulation. The fraud in numbers and text? Not so easy to uncover.
Many journals now require all versions of a gel image that is used in a figure. So, you’d have to fake the full image that is cropped down to the lanes used in the figure. I think there aren’t as many of those raw images around to train AI on… yet.
I predict it will get even worse than that, in the next couple of decades I expect any document or work that has a substantial reward associated with it, either financially or in terms of career advancement or a grade for critical course work in one's major, or penalty such as indictment or conviction, to be backed by a time stamped stack of developing documentation, drafts, revisions, with these time stamp validated against a trusted custodial clock and a seed random string marking the start of work, recorded in some immutable public form.
Accompanying to finish document will be a hash of all of these works along with their associated timestamps, originals of can be verified if necessary to prove a custodial chain of development over a plausible period of time and across multiple iterations of the development of the work - a kind of signed time-lapse slideshow of its genesis from blank page to finished product as if it had a mandatory and global "track changes" flag enabled from the very beginning - by which the entire process can be proved an original human-collaborated work and not an insta-generated AI fiction.
I actually thought that digital timestamps would have been a great use-case for blockchains. They are publicly available and auditable. If you're working from hashes, you don't necessarily need to make the raw data public, just the hash. It is a use-case that has an intrinsic value to the data generator and the future auditor (so you could charge something for it). I know there was some work done on this, but I think it lost momentum due to trying to generate crypto as a value storage medium.
The gold bugs really set back that entire field: the quasi-religious pursuit of “trustless” designs made everything more expensive, but so many problems are far more tractable with trusted third parties both for cost and reduced attack potential because institutional/professional reputations are harder to build than getting n% consensus on a cryptocurrency and don’t have the built-in bug bounty problem.
For example, imagine if university libraries ran storage systems based on Merkle trees with PKI signatures and researchers used those for their papers, code, data inventory (maybe not petabytes of data but the hashes of that data), etc. If there were allegations of misconduct you’d be able to see that whole history establishing that when things were changed and by whom, and someone couldn’t fudge the data without multiple compromised/complicit people in a completely different department (a senior figure can pressure a grad student in their field but they have far less leverage over staff at the library), and since you’re not sharing a database with the entire world you have a much easier time scaling with periodic cross checks (e.g. MIT and Caltech could cross-sign each other’s indexes periodically so you could have confidence that nobody had altered the inventory without storing the actual collection).
Sounds complicated. You could just demand the lab log books. They are supposed to be dated and countersigned. Standard practice is to counter sign outside your group.
The YC company that wanted to sell fake survey results (yes they really had a launch HN with that idea) will surely be the first to sell fake science results next. YC disrupting sciences
Eventually AI will also be able to reliably audit papers and report on fraud.
There may be newer AI methods of fraud, but it will only buy you time. As both progress, committing to record a fraud generated by technology will almost certainly be detected by a later technology.
I would guess that we're within 10 years of being able to automatically audit the majority of papers currently published. That thought must give the authors of fraudulent papers the heebee jeebies.
The problem is that detecting fraud is fundamentally harder than generating plausible fraud. This is because ultimately a very good fraud producer can simply produce output that is identically distributed to non-fraud.
For the same reason, tools that try to detect AI-generated text are ultimately going to lose the arm's race.
It's not a race though. Once the fraud is committed to record it can no longer advance in sophistication. Mechanisms for detection will continue to advance.
I think the argument is that if you produce your fraud from an appropriate probability distribution, any "detection" method other than independently verifying the results is snake oil.
Unfortunately, sometimes someone becomes a bad example. That doesn't make them a "scapegoat", the favored defense of people like that.
A scapegoat is something that takes on all the sins of a lot of others who skate free. If Masliah is the only one who ever suffers, then he IS a scapegoat, but if this article serves to uncover a lot of other bad actors, then he's not. And if his example serves to warn a lot of other scientists to clean up their acts, then his suffering is a benefit.
The language of the article is as low as it is loaded. This is just Derek Lowe covering for the fact that “Science” magazine and the like have let this scoundrel (and many more like him) carry on, without hindrance, for an entire career; pointing the finger anywhere and everywhere but at the journals themselves. None of this is an isolated incident. It is widespread! There is a new scapegoat every month.
> But if the NIH had done that in 2016, they wouldn't be in the position they're in now, would they? How many people do we need to check? How many figures do we have to scrutinize?
It's wrong to think that because there is reports of fraud or systematic error in science you shouldn't trust it. I'm sure all those things exist. But they also exist in every other institution with a lot less self-reflection and self-correction.
Nassim Taleb said that people think weathermen are terrible predictors of the future. He says meteorology is among the most accurate sources of predictions in our lives. But we can easily validate it and we see the mistakes. If we had as much first hand experience with other types of predictions we'd appreciate the accuracy of weatherman. My point is: just because you know the flaws in a system don't assume it isn't better than another.
The Retraction Watch website does a good job of reporting on various cases of retractions and scientific misconduct [1].
Like many others, I hope that a greater focus on reproducibility in academic journals and conferences will help reduce the spread of scientific misconduct and inaccuracy.
I'm the furthest thing from a scientist unless you count 3,000 hours of PBS spacetime, but I love science and so science/academia fraud to me, feels kinda like the worst kinda fraud you can commit. Financial fraud can cause suicides and ruin in lives, sure, but I feel like academic fraud just sets the whole of humanity back? I also feel that through my life I've (maybe wrongly) placed a great deal of respect and trust in scientists, mostly that they understand that their work is of the upmost importance and so the downstream consequences of mucking around are just too grave. Stuff like this seems to bother me more than it rationally should. Are people who commit this type of science fraud just really evil humans? Am I over thinking this? Do scientists go to jail for academic fraud?
I agree with you, science fraud is terrible. It pollutes and breaks the scientific method. Enormous resources are wasted, not just by the fraudster but also by all the other well meaning scientists who base their work on that.
In my experience no, most fraudsters are not evil people, they just follow the incentives and almost non-existent disincentives.
Scientist has become just a job, you find all kinds of people there.
As far as I know no-one goes to jail, worst thing possible (and very rare) is losing the job, most likely just the reputation.
IMO “evil” is a misconception. People have different beliefs and psychological needs, and placed in certain incentive structures that has the outcomes that we see. You can call certain behaviors “evil”, but that doesn’t explain anything about why the behaviors occur.
Nope. “Evil” still provides no explanation and no understanding of why and how things happen there. It’s the same thing as believing in miracles created by a god.
The context here is from the root comment: “Are people who commit this type of science fraud just really evil humans?”. “Just really evil” implies that that there is no other explanation, and that the fraud is committed as a function of them being “really evil”.
I don’t actually know what people mean when they label someone as “evil”, other than “is doing/saying/thinking stuff I find very reprehensible”. Which doesn’t make sense when you insert it into the above statement: “Are people who commit this type of science fraud just humans who do stuff I find really reprehensible?” Well, I guess it sounds like they are.
It seems like people want to assign a character trait when they say “person X is evil”, but I don’t believe such a generic character trait exists (and what exactly it is supposed to mean if it existed). What’s worse, it obfuscates and prevents understanding the actual character traits and circumstances that lead to the respective behavior.
Perhaps if you define evil as a low quantity of ability or commitment to search for and act in accordance to what is ultimately true then that will better resonate with you. Of course, that will necessarily lead to questions regarding the nature of truth and whether it exists, but that is beyond the scope of a short reply :)
I agree, "evil" is a misconception, there exists no such thing as an "evil" person, in reality, just as there is no such thing as a "darling" person, in reality. But both expressions work as an expression of sentiment. When we use it we aim to communicate that we feel no empathy for such people (in case they are "evil"); they can without further ado be thrown in the dungeon. It is a dehumanizing construct enabling hate, same as calling people vermin, or monsters, but with religious connotations, exposing a will to exclude such people from the community (often for god reason), enabling going to war, or to exploit.
However, by removing empathy, we also reduce the possibility to understand the human motivations behind heinous acts (there always are), find solutions, build bridges, make truces, end wars. So maybe we should go lightly on the "evil" stuff, as much as possible.
Physical pain is objective. Someone inflecting physical pain is evil unless it’s in self defense or common sense situations like a doctor performing surgery.
What is a general definition of “evil” that one could derive this from? And how does this relate to the actual reasons why someone would inflect physical pain? Are soldiers in a war evil when they happen to inflict physical pain outside of self defense? Or is that another “common-sense” exception?
The concept is emotionally laden and ill-defined, and has little relation to why the designated behaviors actually happen. It’s an incoherent concept that has no explanatory power.
Exactly. In fact, all things in the universe are subjective except exactly one thing, which is that all other things are subjective. This is epistemological monism, and it's the only coherent view.
Socrates got it. "I know that I know nothing" (else)
I have the displeasure of having acquaintances that have done some pretty bad things, of the fraud and bribery persuasion. They did so because they had no regard of the secondary cosequences. However, this didn't mean 'I understand this horrible secondary consequence is going to happen, but I don't care'. That would be evil. Instead, it's more common to not dedicate an iota of time at thinking of possible negative effects at all.
You'll see this all over risky startups. What starts as hopeful optimism only becomes fraud over time, when the consequences of not committing fraud also seem horrible. It's easy to follow the road until all your choices are horrible in different ways, and they pick the one better for the people around them, yet worse for everyone else.
Our judgment of societal ills and the concept of "evil" rests too much on the question of "is this a bad person?" today. Most people who do heinous things are not bad people, but the fact that they did bad things really ought to be enough to mete out punishment.
Lack of foresight isn't a virtue, it's as much of a vice as knowing the consequences and ignoring them. If you lack foresight and that causes you to commit fraud, you committed fraud, plain and simple. That is evil.
Because we're cowards, and declaring vast swatches of our economy and society to be evil is not good for our future prospects. In other words, you can't tell people not to put radium up their asshole!
It is the same flavor of fraud as financial fraud. It is about personal gain, and avoiding loss.
This kind of fraud happens because scientists are rewarded greatly for coming up with new, publishable, interesting results. They are punished severely for failing to do that.
You could be the department's best professor in terms of teaching, but if you aren't publishing, your job is at risk at many universities.
Scientists in Academia are incentivized to publish papers. If they can take shortcuts, and get away with it, they will. That's the whole problem, that's human nature.
This is why you don't nearly as many industry scientists coming out with fraudulent papers. If Shell's scientists publish a paper, they aren't rewarded for that, if they come up with some efficient new way to refine oil they are rewarded, and they also might publish a paper if they feel like it.
A lot of companies reward employees for publications. Mine certainly does. Also an oil company may not be such a great example since they directly and covertly rewarded scientists for publishing papers undermining climate change research.
It seems like this could ultimately fall under the category of financial fraud, since the allegations are that he may have favorably misrepresented the results of drug trials where he was credited as an inventor of the drug that's now worth hundreds of millions of dollars.
> There was a period of time when science was advanced by the aristocrats who were self funded and self motivated.
From a distance the practice of science in early modern and Enlightenment times might look like the disinterested pursuit of knowledge for its own sake. If you read the detailed history of the times you'll see that the reality was much more messy.
Today we only remember the great thinkers of these times, and tend to see a linear accumulation of knowledge. If you look at the history of the times you realise that at the time there was a vast and confusing babble, it was very hard at the time to distinguish the valid science from the superstition, the blind regurgitation of classical authority, the soothsayers and yes, the fraudsters.
For example Kepler considered his work on the Music of the Spheres (google it) to be more important than, and the ultimate goal of, his research on the mechanics of planetary motion. Newton dabbled in alchemy, and his dispute with Leibnitz was very very bitchy with some dubious jostling for priority. And there was no end of dubious research and outright fraud going on at the time. So no, it was not a golden era of disinterested research.
See for example the wikipedia articles on Phlogiston, The Music of the Spheres, the long and hard fought battle over Epicycles etc
Not the OP, but I remember reading about many twists and turns on the road to various inventions described in Matt Ridley's "How Innovation Works". I personally like "Happy Accidents. Serendipity in Major Medical Breakthroughs in the Twentieth Century" by Morton Meyers.
Pick up an old engineering book at some point, something from mid 1800's or early 1900's and you'll quickly realize that the trust people put on science isn't what it should be. The scientific method works over a long period of time, but to blindly trust a peer review study that just came out, any study, is almost as much faith as religion, specially if you're not a high level researcher in the same field and have spent a good amount of time reading their methodology yourself. If you go to the social sciences then the amount of crock that gets published is incredible.
As a quick example, any book about electricity from the early 1900's will include quite serious sections about the positive effects of electromagnetic radiation (or "EM field therapies"), teach you about different frequencies and modulations for different illnesses and how doctors are applying them. Today these devices are peddled by scammers of the same ilk as the ones that align your shakras with the right stone on your forehead.
Going to need some citations here since the texts that I'm familiar with from that time period are "A Treatise on Electricity and Magnetism" by Maxwell (mid-late 1800s) and "A History of the Theories of Aether and Electricity" by E. T. Whittaker, neither of which mentions anything of the sort. I suspect you are choosing from texts that at the time likely would not have been considered academic or standard.
Your points of memory are not counterpoints.
Those are the ones that lived - and are not indicative of the general quality of science during those times. Obvious survivor bias.
The fact that you can recall those reinforces the point that the value is determined by how long it is useful and remembered, not the fact that it was published.
Indeed, but you are clearly missing the historical context as these were two highly celebrated and referenced texts of the time period by leading scientists. However, it appears that the leading scientific minds (of which Maxwell and Whittaker are) did not include these uses in their texts. I do not dispute that science can be wrong (in fact it is almost always 'wrong' in the end) nor do I dispute that there could have been published research in those applications. I would argue that these applications were likely fringe at best within the scientific community by the mid 1800s.
There are of course incredible scientists that went down disappointing paths (eg Shockley, Dyson, Pauling) in terms of their research output later on, though one must remember that typically this occurs outside their original field of expertise.
If you read my comment you will see that I am asking for the references to the claims the previous author made. I simply provided my own references which werew written at the time and are representative of the times that do not corroborate the tall tale of the previous author. If you have any references to support their claim I would be interested in perusing them.
And what's to say that other highly celebrated and highly referenced texts from that time were not based on bad science or were outright frauds? Your memory of them?
Picking the winners as examples is not good sampling.
The originator explicitly said that 'any engineering book' would contain these references, thus it would seem that this was at least a widespread belief among physicists and engineers at the time. Do you have any example?
Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted. I think that this assertion should be backed by considerable evidence, and that burden is of course mostly on you
I don't dispute that there were doctors applying electricity and/or magnetism to the body in an "un"-rigorous manner, but is there documentation that suggests that the scientists at the time had come to the conclusion that it worked?
Also notably, Whittaker's work was a 'loser'. I chose it specifically for this purpose. I had read parts of it previously because it was a 'loser' as he chose to dispute Einstein's contributions to special relativity.
We've gotten into the territory of just repeating ourselves, so I don't want to do that.
I will say that
> Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted.
Is not correct as far as I believe. Instead, we're saying that there's no reason to believe any study until it has stood the test of time. The longer it remains impactful the better.
I am building from this statement:
> The scientific method works over a long period of time, but to blindly trust a peer review study that just came out, any study, is almost as much faith as religion, specially if you're not a high level researcher in the same field and have spent a good amount of time reading their methodology yourself.
Saying that textbooks from Maxwell's era had misunderstandings and bad information is not saying they are inept, it's saying that that is how it works, it always has and always will be that way. That's it, really. The fact that good science came from it is to be expected, and the fact that bad science existed is also not to be suprising.
I think you interpreted the statement about the 1900s textbooks being wrong as a slander against the entire era, which is not how I read it, and certainly not what I meant to imply by any of my comments.
>Is not correct as far as I believe. Instead, we're saying that there's no reason to believe any study until it has stood the test of time. The longer it remains impactful the better.
Upon rereading I agree, I apologize to you and OP for my misunderstanding. However, ultimately in general I still have to disagree at least semantically with "standing the test of time". I am not really familiar with the processes in biological or social sciences, but from a physical science background, any result of interest will need to be built upon quite quickly. Either some kind of design will be reproduced to improve it or use it, or in the case of a theoretical result it will be awaiting some kind of experiment to validate it.
>> The scientific method works over a long period of time, but to blindly trust a peer review study that just came out, any study, is almost as much faith as religion, specially if you're not a high level researcher in the same field and have spent a good amount of time reading their methodology yourself.
I don't necessarily disagree with this statement (besides the 'long period of time'). Though I would also say that simply mistrusting the result has the same issue, so the only correct way forward seems to me to be to act as if it does not exist until you gain the expertise.
>Saying that textbooks from Maxwell's era had misunderstandings and bad information is not saying they are inept, it's saying that that is how it works, it always has and always will be that way. That's it, really. The fact that good science came from it is to be expected, and the fact that bad science existed is also not to be suprising.
I'm not familiar with textbooks of that era as through personal curiosity I've only read a few. I would still like to see an example of such an occurrence to understand the context under which these treatments are discussed. If these fallacious techniques were widespread enough to be popular in textbooks there must be some kind of literature supporting them?
>I think you interpreted the statement about the 1900s textbooks being wrong as a slander against the entire era, which is not how I read it, and certainly not what I meant to imply by any of my comments.
I will admit to being a bit hotheaded in the initial response, which I apologize for.
You have made so many mistakes in your reading that I would urge you next time to carefully re-read people's posts before responding to them. Also never quote someone without using their actual words. For example, it was not explicitly said that any engineering book would contain those references, it was a general statement not a categorical statement.
>Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted.
No, once again that's not what was said. The concept being communicated is that the scientific method works over long periods of time, not short ones. Over long periods of time, such as 200 years, the work that survives peer review and remains significant today are things like Maxwell's work on electromagnetism as opposed to Dr. Franz Mesmer's work on animal magnetism.
You are taking little bits and pieces of what people are saying, misconstruing them and reinterpreting them, and then forming an argument that is not a genuine representation of the original comment.
>You have made so many mistakes in your reading that I would urge you next time to carefully re-read people's posts before responding to them. Also never quote someone without using their actual words. For example, it was not explicitly said that any engineering book would contain those references, it was a general statement not a categorical statement.
I will admit to some mistakes in comprehension and a poor literal quoting, though I will also maintain that I captured the majority of the essence of what was written in the quote. In the case of "any book about electricity..." vs "any engineering book" the only books that would be relevant to the discussion should be engineering OR science books relating to electricity.
>No, once again that's not what was said. The concept being communicated is that the scientific method works over long periods of time, not short ones. Over long periods of time, such as 200 years, the work that survives peer review and remains significant today are things like Maxwell's work on electromagnetism as opposed to Dr. Franz Mesmer's work on animal magnetism.
With respect to "time" I will refer to another comment I made below in response to the previous OP.
I think it's interesting that you use Mesmer as an example because his work failed to gain the acceptance of the scientific societies of the time, and was in the late 1700s, significantly earlier than the proposed mid 1800s to early 1900s
Because those two texts are the two among literally thousands of scientific publications that have survived the test of time, which is exactly the point being made.
This might seem crazy to hear now, but when Maxwell first published A Dynamical Theory of the Electromagnetic Field in 1865, no one cared, it received very little attention at the time.
It was decades later in 1888 with the work of Hertz that Maxwell's equations started to gain significance within the scientific community.
Here's a more recent (1950) example that I think makes parent's point quite well:
> I assume that the reader is familiar with the idea of extra-sensory perception, and the meaning of the four items of it, viz. telepathy, clairvoyance, precognition and psycho-kinesis. These disturbing phenomena seem to deny all our usual scientific ideas. How we should like to discredit them! Unfortunately the statistical evidence, at least for telepathy, is overwhelming. It is very difficult to rearrange one's ideas so as to fit these new facts in. Once one has accepted them it does not seem a very big step to believe in ghosts and bogies. The idea that our bodies move simply according to the known laws of physics, together with some others not yet discovered but somewhat similar, would be one of the first to go.
Anyone on this site who doesn't know what this is from should feel a bit of shame in the current era of hype around machine intelligence, so I'll leave it as an exercise to the reader if you aren't already familiar with this paper.
Alan Turing was an incredible computer scientist and mathematician.
Unfortunately he is out of his area of expertise in physics and human biology/neuroscience (? not sure where telepathy would be if it was to be rigorously studied). This is akin to Freeman Dyson on global warming.
That scientists can have strange ideas is something nobody can dispute. That those strange ideas enter into scientific legitimacy is another story entirely.
The point is that I don't believe Turing's ideas were widely considered strange at the time. The point is more that, even under conditions of honest actions, it's very easy for educated, smart, and sincere thinkers to take for fact something that with time we believe is wildly not fact.
Science even at it's most sincere should always be approached with thoughtful skepticism. The phrase that I hear touted often these days "trust the science", is in essence not how science should be thought of.
There is a difference between "not considered strange at the time" and "science" via scientific publication and subsequent consensus validating the idea. I have mentioned that luminaries can have odd ideas multiple times in this thread, it's not something I seek to deny. However, as I continue to reiterate, these ideas are generally:
1. outside their areas of expertise
2. not validated by independent scientific research
I completely agree that science should be approached with thoughtful skepticism, and I agree that 'trust the science' might not necessarily be the best semantics to use. However, it is not clear that skepticism by all parties should be considered with equal weight. Most of the times, people should "trust the science" because they are not equipped to be skeptical.
Sure, check out Audel's Electric guide, 1927 print, volume 10, for example. This is a series of books intended for practitioners at the time, in wide circulation in the USA. The "Electric Therapeuthics" section changed a lot with each edition so it's even interesting to compare across older versions if you can find multiple. I didn't want to reference specific old niche books but if you're willing to ebay it, I guess you can check. My point was more general than just electric engineering though.
We use EM radiation for illnesses and doctors apply them. It's one of the most important diagnostic and treatment options we have. I think what you're referring to is invalid therapies ("woo" or snake oil or just plain ignorance/greed) but it's hard to distinguish those from legitimate therapies at times.
Gamma knife? Basically the entire field of radiotherapy?
TMS is magnetic, not EM (the coil generates a magnetic field, which induces localized currents in the body being treated)
I can't parse what you are saying, but there's a difference between EM radiation and a magnetic field (and the resulting locally induced currents).
Think in terms of an MRI machine: it puts you in a giant magnet (causing the various nuclear spins to align with the field) and then sends a bunch of EM radiation (radiofrequency). The former is a magnetic field, not EM radiation.
OP is making a distinction between "EM Radiation" (i.e. "light") and "Quasistatic fields".
This is warranted because they are pretty different - light has a frequency distribution, diffracts, etc, and can be focused to propagate energy over distances large compared to its source, whereas quasistatic fields (by definition) have no frequency distribution, and die as 1/r^2 or faster
the best example is psychology. the entire field needs to be scrapped and started over, nothing you read on any of those papers can be trusted, it's just heaping piles of bad research dressed with a thin veil of statistical respectability.
I treat the field like medicine in 19th century. Nice motive, but unless someone drag me to operating room because I've just been shot, I won't have any surgery.
Exactly, but the problem is that everyone has to be the first to publish about something and then every paper after that which deals with the same thing is relegated to second-level venues (or lower). So whatever is the first word about something becomes the accepted and is super hard to contest.
The default state of the human brain almost seems to be a form of anti-science, blind faith in what you already believe, especially if you stand to gain personally from what you believe being true.
What is most incredible to me is even knowing and believing the above, I fall prey to this all the time.
I think the error is putting trust in scientists as people, instead of putting trust in science as a methodology. The methodology is designed to rely on trusting a process, not trusting individuals, to arrive at the truth.
I guess it also reinforces the supreme importance of reproducibility. Seems like no research result should be taken seriously until at least one other scientist or group of scientists are able to reproduce the result.
And if the work isn't sufficiently defined to the point of being reproducible, it should be considered a garbage study.
There is no way to do any kind of science without putting trust in people. Science is not the universe as it is presented. Science is the human interpretation of observation. People are who carry out and interpret experiments. There is no set of methodology you can adopt that will ever change that. "Reproducibility" is important, but it is not a silver bullet. You cannot run any experiment exactly in the same way ever.
If you have independent measurements you cannot rule out bias from prior results. Look at the error bars here on published values of the electron charge and tell me that methodology or reproducibility shored up the result. https://hsm.stackexchange.com/questions/264/timeline-of-meas...
TFA is about a person who literally faked the observations. Everyone on this sub is trying to shoehorn in their preferred view of "how to fix science" when the problem here has nothing to do with any of it.
The initial GP comment made the point that, at some level, science requires trust. And (in the case of TFA) specifically trust that the person making the observations is actually performing the experiments and recording them correctly -- rather than making them up. You can verify and replicate (and we do quite a bit of that, modulo the fact that resource constraints are a huge problem in science) but without some degree of trust you're in trouble.
But the OP's suggestion was that to fix this specific problem of faked observations, you should separate interpretation and observation. I don't see how that fixes this problem at all. And in my view the first step in solving the problem is to come up with some sense of how serious the problem is: meaning, rather than dwelling on each terrible isolated case and panicking, try to determine what the actual prevalance is and what the overall impacts are. With that information you can make resource-allocation decisions in how to address it. The HN response is much too emotional for anything useful to come of it (except for more anger and confirmation bias.)
You're talking about science, the methodology. They're talking about science, the social institution. Scientists lying is a problem with the latter, not the former.
This error really went viral during the pandemic and continues to this day. We're in for an Orwellian future if the public does not cultivate some skeptic impulse.
I'd say the public needs to develop some rational impulse, it already has plenty of skepitism to the point where people no longer trust science the methodology. Instead, they genuinely believe there is some alternative to finding the truth, and now simply believe the same old superstitions and bunk that people have prior to the scientific revolution.
Speaking of Orwell, I don't think science comes into it. Rather, when people stop believing in democracy, things will degenerate into authoritarianism. It's generally pretty hard to use science the methodology to implement an authoritarian government as the scientific method by definition will follow the evidence, not the will of a dictator.
However, something that looks like science but isn't could be used, especially if the public doesn't understand science and thus can't spot things that claim to be science but don't actually follow the scientific method.
> I'd say the public needs to develop some rational impulse, it already has plenty of skepitism to the point where people no longer trust science the methodology.
Methodologies are inanimate - I may trust that a methodology is fine, but once Humans become involved I do not trust.
> Instead, they genuinely believe there is some alternative to finding the truth
There are several alternate means, the field of philosophy (that birthed science) has been working on such problems for ages, and has all sorts of utility, just sitting there waiting to be used by Humanity.
> and now simply believe the same old superstitions and bunk that people have prior to the scientific revolution.
Not possible for you to know, unless there are indeed forms of supernatural (beyond current scientific knowledge) forms of perception.
> Rather, when people stop believing in democracy, things will degenerate into authoritarianism.
Once again, not possible for you to know.
> It's generally pretty hard to use science the methodology to implement an authoritarian government
COVID demonstrated that to be incorrect.
> as the scientific method by definition will follow the evidence, not the will of a dictator.
Incorrect. Something defined to be true necessarily being true only works in metaphysics, such as linguistics.
And again, the scientific method is inanimate.
> However, something that looks like science but isn't could be used, especially if the public doesn't understand science and thus can't spot things that claim to be science but don't actually follow the scientific method.
On a scale of 1 to 10, how comprehensively and accurately do you believe you understand science?
Critical thinking = the ability to be skeptical, literally it is the ability to criticize.
Great critical thinkers become lawyers, post modernist intellectuals, and other parts of the "talking" class of intellectuals. Unfortunately, it's far easier to talk shit than it is to build things. We've massively over-valued critical thinking over constructive thinking.
Most people want to dunk on science. Few people want to submit their own papers to conferences. Many people act like submitting papers is impossible for non-Ph.D's. We have a lack of constructive oriented thinking.
>Many people act like submitting papers is impossible for non-Ph.D's.
I agree. But academia reinforces this perception. I feel like PhD's only give serious consideration to the utterances of other PhD's. The rest of the public consists of the unwashed masses, and at best gets the smiling-nod treatment from teh PhD.
PS I (a non-PhD) managed to publish a paper during the pandemic (doi: 10.3389/fphar.2022.909945 ). One of the biggest barriers was the item you mentioned quoted above, and the bogeyman of "epistemic trespass" in general, as operating in my own psychology. I've since become noisy in advocating for the #DeSci movement.
How does this work for things like COVID vaccines, where waiting for a reproduction study would leave hundreds of thousands dead? Ultimately there needs to be some level of trust in scientific institutions as well. I do think placing higher value on reproducibility studies might help the issue somewhat, but I think there also needs to be a larger culture shift of accountability and a higher purpose than profit.
I believe if we taught philosophy in school to a non-trivial level we wouldn't have to rely on trust/faith.
I wonder if it's possible to get people to wonder why the one discipline that has the tools to deal with all of these epistemic, logical, etc issues isn't taught in school. You'd think it would be something that people would naturally wonder about, but maybe our fundamentalist (and false) focus on science as the one and only source of knowledge has damaged our ability to wonder independently.
Suppose you need to make a decision on a topic that's contingent on P being true, which someone has already tested. How would you go about making the decision without testing P yourself (because that would mean that you would have to do the same for every decision in your life)?
> How would you go about making the decision without testing P yourself (because that would mean that you would have to do the same for every decision in your life)?
This does not seem necessary for my ask above.
There may be many approaches, some impossible/invalid, but perhaps not all.
The way I sum it up is: science is a method, which is not equivalent to the institution of science, and because that institution is run by humans it will contain and perpetrate all the ills of any human group.
Science is an anarchic enterprise. There is no "one scientific method", and anyone telling you there is has something to sell to you (likely academic careerism). https://en.wikipedia.org/wiki/Against_Method
You're far from a scientist, so it's easy for you to put scientists/academia on a pedestal.
For most of the people who end up in these scandals, this is just the day job that their various choices and random chance led up to. they're just ordinary humans responding to ordinary incentives in light of whatever consequences and risks they may or may not have considered.
Other careers, like teaching, medicine, and engineering have similar problems.
It's complicated. Historically scientific fraud could be construed as 'good-intentioned' - typically a researcher in a cutting edge field might think they understood how a system worked, and wanting to be first to publish for reasons of career advancement, would cook up data so they could get their paper into print before anyone else.
Indeed, I believe many academic careers were kicked off in this manner. Where it all goes wrong is when other more diligent researchers fail to reproduce said fraudulent research - this is what brought down famous fraudster Jan Hendrik Schön in the field of plastic-based organic electronics, which involved something like 9 papers in Science and Nature. There are good books and documentaries on that one. This will only be getting worse with AI data generation, as most of those frauds were detected by banal data replication, obvious cuts and pastes, etc.
However, when you add a big financial driver, things really go off the rails. A new pharmaceutical brings investors sniffing for a big payout, and cooking data to make the patentable 'discovery' look better than it is is a strong incentive to commit egregious fraud. Bug-eyed greed makes people do foolish things.
People like us think scientists care about big-money things, but they largely don't care about that stuff as much as they care about prestige in their field. Prominent scientists get huge rewards of power and influence, as well as indirect money from leveraging that influence. When you start to think that way, the incentives for fraud become very "minor" and "petty" compared to what you are thinking of.
As a collective endeavor to seek out higher truth, maybe some amount of fraud is necessary to train the immune system of the collective body, so to speak, so that it's more resilient in the long-term. But too much fraud, I agree, could tip into mistrust of the entire system. My fear is that AI further exacerbates this problem, and only AI itself can handle wading through the resulting volume of junk science output.
In my view, prosecuting the bad actors alone will not fix science. Science is by its own nature a community because only a small number of people have the expertise (and university positions) to participate. A healthy scientific discipline and a healthy community are the same thing. Just like the "tough on crime" initiative alone often does not help a problematic community, just punish scientific fraud harshly will not fix the problem. Because the community is small, to catch the bad actors, you will either have insiders policing themselves, or have an non-expert outsiders rendering judgements. It's easy for well-intention-ed policing effort to turn into power struggles.
This is why I think the most effective way is to empower good actors. Ensure open debate, limit the power of individuals, and prevent over concentration of power in a small group. These efforts are harder to implement than you think because they run against our desire to have scientific superstars and celebrities, but I think they will go a long way towards building a healthy community.
I also watched almost all episodes of PBS Spacetime. Some of them multiple times. I'm so happy that Spacetime exists and also that Matt was recruited as a host (in place of Gabe). Highly recommended channel, superb content!
Evil is a much simpler explanation than recognizing that if you were in the same position with the same incentives, you would do the same thing. It's not just one event, it's a whole career of normalizing deviation from your values. Maybe you think you'd have morals that would have stopped you, maybe those same morals would have ensured you were never in a position to PI research like that.
Scientific fraud can also compound really badly because people will try to replicate it, and the easiest results to fake are usually the most expensive...
This is pretty funny. I usually hear this kind of language when a religious person is so devastated when their priest or pastor does something wrong that it causes them to leave their religion altogether. Are you going to do the same thing for scientism?
I'm not a particularly religious person, I didn't realize what you described is something that happens with any great frequency. Never the less, I suppose one is able to leave a particular place of worship and not leave a religion, as it is with any way people form their views on something societal like this, it's on a spectrum? Religion, Politics, Science, Sex, Education, whatever.
Generally, the fields that have a Nobel in them attract the glory hounds and therefore the fraudsters. The ones that don't, like geology or archeology for example, don't get the glory hounds.
Anytime you see champagne bottles up on a professor's top shelf with little tags for Nature publications (or something like that), then you know they are a glory hound.
When you see beer bottles in the trash, then you know they're in it for more than themselves.
As a scientist, I agree, although for not quite the reason you gave. Scientists are given tremendous freedom and resources by society (public dollars, but also private dollars like at my industry research lab). I think scientists have a corresponding higher duty for honesty.
Jobs at top institutions are worth much more than their nominal salary, as evidenced by how much those people could be making in the private sector. (They are compensated mostly in freedom and intellectual stimulation.) Unambiguously faking data, which is the sort of thing a bad actor might do to get a top job, should be considered at least as bad a moral transgression as stealing hundreds of thousands or perhaps a few million dollars.
(What is the downside? I have never once heard a researcher express feeling threatened or wary of being falsely/unjustly accused of fraud.)
> Stuff like this seems to bother me more than it rationally should.
It's bothering you a rational amount, actually. These people have done serious damage to lots of lives and humanity in general. Society as a whole has at least as much interest in punishing them as it does for financial fraudsters. They should burn.
In the future those who commit fraud are not likely leave trace in Western blot and photomicrograph audit.
When the experiments are significant, double blind is not enough. You need external auditors when conducting experiments. Preferably separate team making experiments from those who design them.
For all the complaints about AI generated content showing up in scientific journals, I'm exited for the flip side, where an LLM can review massive quantities of scientific publications for inaccuracies/fraud.
Ex: Finding when the exact same image appears in multiple publications, but with different captions/conclusions.
The evidence in this case came from one individual willing to volunteer hundreds of hours producing a side by side of all the reports. But clearly that doesn't scale.
Wouldn’t it be cool if people got credit for reproducing other people’s work instead of only novel things. It’s like having someone on your team that loves maintaining but not feature building.
> For all the complaints about AI generated content showing up in scientific journals, I'm exited for the flip side, where an LLM can review massive quantities of scientific publications for inaccuracies/fraud.
How would this work? AI can't even detect AI generated content reliably.
Not in a zero shot approach. But LLMs are more than capable of solving a similar scenario to the one presented:
- Parse all papers you want to audit
- Extract images (non AI)
- Diff images (non AI)
- Pull captions / related text near each image (LLM)
- For each image > 99% similarity, use LLM to classify if conclusions are different (i.e. highly_similar, similar, highly_dissimilar).
Then aggregate the results. It wouldn't prove fraud, but could definitely highlight areas for review. i.e. "This chart was used in 5 different papers with dissimilar conclusions"
LLMs might find some specific indications of possible fraud, but then fraudsters would just learn to avoid those. LLMs won’t be able to detect when a study or experiment isn’t reproducible.
Of course, but increasing the difficulty of committing fraud is still good. Fraudsters learn to bypass captchas as well, but they still block a ton of bad traffic.
Won't the scientist use some relatively secure/private model to fraud-check their own work before submitting? If it catches something, they would just improve the fraud.
I'm hoping it won't have the same results as AI Detectors for schoolwork, which have marked many legitimate papers as fraud, ruining several students' lives in the process. One even marked the U.S. Constitution as written by AI [1].
It's fraud all the way down, where even the fraud detectors are fraudulent. Similar story to the anti-malware industry, where software bugs in security software like CrowdStrike, Sophos, or Norton cause more damage than the threats they prevent against.
If you are familiar with academia you'll realize the academic dishonesty policy is essentially the playbook by which academics behave. The author is surprised that Eliezer Masliah purportedly had instances of fraud spanning 25 years. I bet the author would be even more surprised to find out that most academics are like that for the entire duration of their career. My favorite instance is Shing-Tung Yau, who is still a Harvard professor, who attempted to steal Grigori Perelman's proof of Poincare's conjecture (a millenium prize problem <https://www.claymath.org/millennium-problems/> that comes with a $1MM prize and $10k/mo for the rest of one's life; Perelman rejected all of it.)
I mean, get this: an extremely gifted Mathematician living on a measly salary in Russia had had his millenium prize almost stolen by a Harvard professor. What more evidence do you need?
From personal experience, it is all I've seen. Could anyone be in a position to extrapolate to all of academia without speaking from personal experience? I'm not speaking of all academics (hence 'most'). It's a statement similar to "Hollywood has a drug problem" or something of that sort.
My advice to anyone going into Hollywood would be to stay away from drugs; my advice to anyone going into academia is to treat every interaction as if you've just sat at a poker table in Las Vegas.
I work in Hollywood. I am not sure it has more of a drug problem more than say tech or finance. Maybe it does-- I don't know. The point is when a celebrity is a drug addict you hear about it. When a banker or a lawyer is you don't.
Our experience of things has a lot of bias toward what we want to hear. Generalization plays into sterotypes and ideology.
I believe that tech and finance also have a drug problem. Those that sell expensive drugs like cocaine go after rich clients. You work in Hollywood, but have you been attending wild private parties? I've worked in academia and I was in the thick of it, I've experienced first hand the fraud I'm talking about, and it was a large part of my experience, not some side note. Perhaps it's an uncomfortable truth that academia is in the state it is in, but again, it is of utmost importance to warn younger people to its perils. (Act as if you're at a poker table at all times.) In any case, how do you know that it isn't your biases that prevent you from considering what I describe? What is so surprising with the claim that people who are very incentivized to steal and commit fraud do so if they are not punished for it?
edit: and it's not things I've heard, instead it is direct experiences, i.e. people stole my work, and things like that. As a graduate student to watch professors come to you with problem X, take what you've said (an actual solution) and publish a paper without attribute, that sort of thing; to report it and have nothing be done about it, et cetera, and on it goes, it's just instance after instance of such behavior, or the million ways in which they are careful to trick you into working on their problems without receiving attribute. One such trick for example, that again happened to me, is that after a conference talk I got into an e-mail discussion where I explained my approach; I was told that "they already have these results" (the trick here was to divulge less in the talk than what was currently known in order to be able to avoid "significant progress by another person" in the case another person does share new progress that they have already established, and hence not having to share attribution.) It turned out that our discussion was enough for them to go from n=3,4 to a general formula involving primes, because I pointed out a certain property they had not noticed. This is just a single example of the sorts of tricks, aside from total fraud, that happen, and one of the milder incidents I had happen to me.
I extrapolate to all of academia, but not to all academics (persons working in academia). My methodology is based on my intuition and my experiences. Already in this YC article the comments appear to be akin to the first meeting between battered housewives. You don't have to believe me or others, I'm just issuing a warning to anyone thinking of getting into academia: be alarmed and alert, and always careful. It's nothing like the movies portray academia to be, instead it's a thieves den, or a poker table, etc, you get the point.
I'm a recovering academic, and have not published since not long after defending my dissertation.
I blame this behavior entirely on "publish or perish". The demands for novel, thoughtful and statistically-significant findings is tremendous in academe, and this is the result: cheating.
I left professional academia because I resented the grind, and the push to publish ANYTHING (even reframing and recombining the same data umpteen times in different publications) in an effort to earn grants or attain tenure.
The academia system is broken, and it cannot be repaired with minor edits, in my opinion. This is a tear out and do over scenario for the academic culture, I'm afraid.
If there's this much overt, deliberate fraud and dishonesty in all of our research institutions, the quantities of soft lying and fudging are inconceivable.
We need to seriously rethink our approaching to stewarding these institutions and ideas, public trust is rightfully plummeting.
Is there no liability for the author? There are billions of dollars wasted in drug trials and research that can be tied to this fraud. Surely they can face some legal issues due to this?
Like all things in life that have risks of fraud, negligence or potential failure, insurance could be the answer.
Want to publish in a peer reviewed paper? Well then your institution or you should take out a bond or insurance policy that guarantees your work is accurate. The insurance amount would fluctuate based on how big of impact this study could have. Is it a drug that will be consumed by millions? Big insurance policy. Is it a behavioral study without much risk... small insurance policy.
Is a a person in an institution found caught committing fraud, well now then all papers from that institution now have higher premiums.
Did you sign off on a peer reviewed paper that was fraud? Well now your premiums are going up also.
Insurance costs too high to publish? Well then keep doing research until the underwriters are satisfied that your work isn't fraud and adjust the premiums down.
It adds a direct near-term economic incentive to publish honestly and punishes those that abuse the system.
In other words, you are suggesting more stringent peer review conducted by insurance companies. And because insurance companies are too small to have sufficient in-house expertise on every topic, the reviews will be usually done by external consultants. The costs might be from $10k for simple papers to hundreds of thousands for large complex papers.
The insurance model does not really work when the cost of evaluating the risks far outweighs the expected risks.
That is like saying my insurance company has to follow me around for a week while I drive before they can underwrite a policy. If there is money to be made, and money to be lost, the actuaries will find a way.
The problem could be, that it may become impossible to publish certain kinds of papers that are very well supported and valuable because no institution can afford the insurance.
> that it may become impossible to publish certain kinds of papers that are very well supported and valuable because no institution can afford the insurance.
What type of research would that be? Just publish it online without insurance and everyone will treat it as it unverified and uninsured... separate from other research that is.
Once the risk of the publishing research has gone down (i.e. reputable peers approve, or the findings were replicated), the cost of the insurance goes down also.
if something is so costly to insure, there would be a reason and thus the system works.
If it is possible to advance your career by publishing uninsured research then we've just renamed the problem, although I do like the idea of adding this structure. Eventually there could be so much of it that it would become an accepted norm that your research isn't actually published in a journal until five years after you informally publish it. Other scientists in the field have to be abreast of the latest findings, so now these informal publications are the true journals.
I see your point, the success of this would have to align with a change in the broader academia to only cite research from insured researchers.
The "organic" way this would happen is if there was a shift so that journals with insured research are far more valuable than uninsured research. Or perhaps if companies started suing researchers for negligence and fraud and recuperate costs if they used research that was later proved to be fraud.
In the literary world, anyone can publish a book, but a book from o'reilly caries with it a different level of authority and diligence then a self published book or blog post.
So the shift would have to be that your career can't advance without publishing a bonded and insured paper.
But that is not how research works in Academia. They have to follow the bleeding edge of the field, or they may be doing work of their own that is already irrelevant. They will not wait until a consortium of insurance companies and underwriters have done the actuarial analysis and come up with an underwriting product that the institution has funded (and what is the institution's business model for recovering this cost in a field of pure research, anyway?)
You are not the first person in the world to own a home or drive a car. Insurance companies can offer you cost-effective insurance, because you are doing effectively the same things as many other people.
Science is largely about doing novel things and often being the first person in the world to try something. In order to understand the risks, you have to understand the actual research, as well as the personalities and personal lives of the people doing it.
Then there is the question of perverse incentives. Research fraud is not a random event but an intentional action by the people who take the insurance. If they manage to convince you to underwrite their research, they know that the consequences of getting caught will be less severe than without the insurance, making fraud more likely. Normally intentional fraud would not be covered by the policy, but here covering it would be the explicit purpose of the insurance.
Insurance companies insure one off events all the time. You can literally insure anything, its just a matter if the premiums outweigh what you perceive as the risk. "Uninsurable" just means the price is too high to be considered practical.
The research might be novel, but the procedures for research and publication are very similar. So insurance companies would just make sure that you followed a protocol which minimizes their risk.
perverse incentives are taken into account by insurance. Insuring someone is always a adversarial back and forth to determine if they are being truthful or not. Which is why Life insurance companies require a physical. They don't just have you self report and then accept it as fact.
Industry professionals like lawyers and doctors carry malpractice insurance. A lawyer can still commit fraud. Insurance isn't a black and white thing. It is a sliding scale that ties risk to a monetary value.
Its not rocket science. Just actuarial science. ;)
> The research might be novel, but the procedures for research and publication are very similar.
This is wrong.
Some time ago, I completed the checklists for publishing a paper in a somewhat prestigious multidisciplinary journal. Large parts of the lists were about complying with various best practices and formal requirements in different fields. I often didn't even understand the questions outside my field. And the questions nominally within my field were often category errors. They assumed a mode of doing research that was far from universal. Overall, the process was more frustrating than (let's say) applying for a US visa.
I think you are desperately trying to fit something black and white rather than thinking critically that there is a spectrum of research, some of which is similar to others which can easily have procedures for insuring and others that are more complex that require more diligence from the insurance company. Just like nearly every single thing an insurance company does.
Yes there is novel research that has never been done before? So what? That doesn't change if you can get insurance or not. Thats a failed argument from the beginning.
Anyways you don't seem to be having a discussion in earnest and instead you seem to be intentionally disregarding large pieces of the above arguments and trying to shoehorn in your idea that if there is unique research being done that it means that it is impossible to tell the risk of anything. Kinda silly.
The cases that would require more diligence from the insurance company are the kind of research that should be encouraged. Breakthroughs are more likely to happen when people take risks and try something fundamentally new, instead of adhering to the established forms. Your insurance model would discourage such research by making it more expensive.
Additionally, even if we assume that the insurance model is a good idea, it should be tied to individual researchers, not universities. The entire model of university research is based on loose networks of independent professionals nominally employed by various organizations. Universities don't do research, they don't own or control the projects, and they don't have the expertise to evaluate research. They are just teaching / administrative organizations that provide services in exchange for grant overheads.
> you are suggesting more stringent peer review conducted by insurance companies
Absolutely not. Underwriters are smart. They use other variables and methods for determining risk. They don't need to directly recreate and peer review the research themselves.
Not only are there billions of dollars wasted, there are many, many lives wasted. If the billions had gone in a direction that was actually promising, maybe there would be treatments that would have saved millions of person-years of quality lifetime. This person is basically a mass-murderer.
I was thinking about it: If I come across someone seriously injured, try to help them, and accidentally hurt them, I'm protected (in many places) by Good Samaritan laws.
But if a health care professional does the same thing, and does something negligent, then they are usually liable. They are professionals and are held to a different standard. Similarly, that's why lawyers keep writing: this is not legal advice and you are not my client.
Perhaps a professional in science should have higher standards. Obviously they shouldn't be sued for being wrong - that would destroy science, disregard the scientific method's means to address inaccuracy, and go against science's nature as the means to develop new knowledge. But intentionally deceiving people perhaps should be illegal and/or create liability: When you publish something, people depend on its fundamental honesty and will act on it.
The US has the Office for Research Integrity which can prosecute scientific fraud cases, but it only does a handful of cases per year.
To put the scale of this problem in perspective, the ORI was set up in the 1970s after Congress became concerned at widespread reports of scientific fraud. It clearly didn't work, but hangs around regardless.
It's ultimately a culture problem. Until academics have the same level of respect as ordinary corporate employees, you're going to get judges and juries who let them off scott free.
The line between outright fraud, bad methods correctly implemented, messy data, and implementation bugs is fuzzy. Trying to criminalize anything not very very clearly #1 quickly turns into a case of “show me the man and I’ll show you the crime”. You think groupthink in academia is bad just wait until professional disputes lead to jail time for the loser.
The fact that some areas are gray shouldn't prevent us from demanding legal consequences when the fraud is gross and deliberate, as appears to be the case here.
Once, at 3Com, Bob Metcalfe introduced a talk by one of his MIT professors with the little joke, "The reason academic politics is so vicious is that nothing's at stake."
The guy said, "That depends on whether you consider reputation 'nothing.' "
I guess what that shows is, you can always negotiate and compromise over money, but reputation is more of a binary. An academic can fake some work, and as long as he's never called on it, his reputation is set.
So yeah, a little more fear of having one's reputation ruined would go a long way towards fixing science.
I have always said that while professors get paid less money than in industry, they are compensated in reputation to make up for it. Status and reputation are the currency of academia.
But this is really a societal/political issue: since we decided that economic capital is king and symbolic capital not that much… (This is really the story of the last four decades or so.)
Well, this is about Pierre Bourdieu, and he had a few things to say about academia, as in Homo Academicus.
And I'm not sure what example could illustrate the problem with the lopsided valuation of economic capital and the general devaluation of symbolic capital (as compared to pre-1980s, we have since undergone a social revolution of considerable dimensions, which is also why there isn't an easy fix) better than this one.
Socio-economic issues aren't one-dimensional, in fact they're very complex. Most of our systems and beliefs are socially constructed.
Humans are, by our biology, social creatures. Modern humanity more than ever before. If you're not considering the social effects, then IMO you're not addressing anything of value.
Not many people in the academic/technical people realize this, often for their entire lives. In their naive worldview, they cannot even imagine that people can stoop that low.
(embarrassingly and shamefully I used to be one of those naive people)
The problem being, we have "economized" academia, by things like "publish or perish", a citation pseudo stock market or third party funding, and all incentives are built around this pseudo-economy. Which also imports all the common incentives found in economy…
A caveat that "reputation", like competence, is more variagated and localized than is often appreciated. As with someone who is highly competent and well regarded in their own subfield, while simultaneously rather nutter about some nearby subfield where they don't actually work.
One can have a reputation like "good, but beware they have a thing for <mechanism X>". Or "ignore their results using <technique> - they see what they want to see". Subtext that gets passed among professors chatting at conferences, and to some extent to their students, but otherwise isn't highly accessible.
When people speak of creating research AI's using just papers... that's missing seemingly important backchannels. And corresponding with authors. Attempting research AI as developing-world professionally-isolated professor.
Universities became tax funded and the consequences is warm bodies filling chairs. I have experience with a number of big name unis in the U.S. they are all about office and national politics. It's not about the work and hasn't been for a while now.
Defund universities. No more student loans, make them have to earn their place in the market or we will continue to suffer under the manipulated system that is actually killing students.
> Defund universities. No more student loans, make them have to earn their place in the market or we will continue to suffer under the manipulated system that is actually killing students.
This... it's no longer about value its about optics... Problem exists in most industries now. The pendulum needs to swing back the other way before it's too late to stop the decay...
I had a feeling academia was just run a ran by people letting blatant fraud, exploitation and abuse of phd students, stealing during peer-review, and just other forms of plagiarism, fraud, and exploitation slide by. They let it slide by because correcting these things would lead to massive changes in academia that might put them out of jobs.
Every year that feeling becomes more certain. Glad I quit the track in grad school.
I feel terribly for all the incredibly smart and hard working academics that remain honest and try to make it work. They do what they love, otherwise they wouldn't do such intensive work with so much sacrifice.
It is really disheartening too because academia only turns on the "honesty filter" when it comes to minor grad students that pissed off the wrong people. But you can do all this fraud constantly and become president of harvard if you know the right politics.
Dishonest lot. I hope karma is real so they get what is coming to them for taking advantage of people that just love to increase humanity's knowledge.
You're being downvoted because you're correct—HN is an eco chamber for zealous regurgitation of opinions of the academy and media—institutions that have decayed. It's been happening slowly for awhile, but now things are starting to come apart at the seams.
It is really annoying because a common response is
"We know academia is bad. But this is the best we have and it is hard to improve"
when that is false on two counts.
1. If you had said the same thing before 2016 or covid, people would not agree that academia is rife with fraud or worthy of skepticism.
2. The same people saying the system dismiss how can it be improved are the same ones that would suffer from disruption as you say. They have the power to dismiss these arguments to begin with.
When I hear someone say, 'We know academia is flawed, but it's the best we have, and it's hard to improve,' I can't help but feel a deep, seething frustration.
It's profoundly insulting and grotesque—on par with excusing the inexcusable.
Accepting this degree of mediocrity is as repulsive as tolerating the most heinous acts imaginable. I've confronted people directly with this, to their face, because to me, it's inconceivable how anyone can be okay with such a vile acceptance of the status quo.
If society was even slightly capable of rational action . . . (legally, I cannot complete this sentence).
You are being downvoted because you extrapolating from one fraud case to call all scientists dishonest.
I can do it too. Person named SpaceManNabs made a bad post. There for all posts by SpaceManNabs, and probably all posts on HackerNews are bad. A dishonest lot.
> from one fraud case to call all scientists dishonest
I specifically mention that the majority of scientists are not dishonest. The majority of scientists are not running academia. The majority of scientists are suffering from this system, to differing degrees.
If I were as rude as you, I'd extrapolate on reading ability, especially since it is not just one fraud case.
Regardless, even if I was wrong on that, all my other criticisms of academia still stand, like exploitation of the phd students. I really hope the grad student unions get what they want.
I appreciate your response though. Makes me feel confident that it is just salty people on HN that hate truth, because otherwise, why would you mischaracterize what I said?
Anecdotally, during my (fairly short-lived) academic career, in which I did research with three different groups, 2/3 of them were engaging in fraudulent research practices. Unfortunately the one solid researcher I worked for was in a field I wasn't all that interested in continuing in, and as a naive young person who believed in the myth of academic freedom and didn't really understand the funding issue, I jumped ship to another field, and found myself in a cesspool of data manipulation, inflated claims, and all manner of dishonest skullduggery.
It all comes down to lab notebooks and data policies. If there is no system for archiving detailed records of experimental work, if data is recorded with pencils so it can later be erased and changed, if the PI isn't in the habit of regularly auditing the world of grad students and postdocs with an eye on rigor and reproduciblity, then you should turn around and walk out the door immediately.
As to why this situation has arisen, I think the corporatization of American academics is at fault. If a biomedical researcher can float a false claim for a few years, they can spin their research off to a startup and then sell that startup to a big pharmaceutical conglomerate. If it fails to pan out in further clinical trials, well, that's life. Cooking the data to make it look attractive to an investor - in the almost completely unregulated academic environment - is a game that many bright-eyed eager beavers are currently playing.
As supporting evidence, look at mathematical and astronomical research, the most fraud-free areas of academics. There's no money to be made in studying things like galactic collisions or exoplanets, the data is all in the public domain (eventually), and with mathematics, you can't really cook up fraudulent proofs that will stand the test of time.
> mathematical and astronomical research, the most fraud-free areas of academics. There's no money to be made
So we're systemically safeguarding the quality of astronomy research, by setting up a gradient (at MIT: restaurant catering for business talks, pizza for CS, stale cookies for astronomy) to draw off some flavors of participants and thus concentrate others?
As a scientist who has published in the neuroscience space, I don’t what to say other than the incentives in academia are all messed up. Back in the late 90s, NIH made a big push on ‘translational research”, that is, researchers were strongly encouraged to demonstrate their research had immediate, real world benefits or applications. Basic research and the careful, plodding research needed to nail down and really answer a narrow question was discouraged as academic navel-gazing.
On one hand, it seems the push for immediate real world relevance is a good thing. We fund research in order that society will benefit, correct? On the other hand, since publications and ultimately funding decisions are based on demonstrating real world relevance, it’s little surprise scientists are now highly incentivized to hype their research, p-hack their results, or in rare cases, commit outright fraud in an attempt to demonstrate this relevance.
Doing research that has immediate translational benefits is a tall order. As a scientist you might accomplish this feat a few times in your career if you’re lucky. The rest of the corpus of your work should consist of the careful, mundane research the actual translational research will be based upon. Unfortunately it’s hard to get that foundational, basic, research published and funded nowadays, hence the messed-up incentives.
There's evidence that the turning point was in the 90s but I suspect the real underlying problem is indirect funds as a revenue stream for universities, combined with the imposition of a for-profit business model expectation from politicians at the state and other levels. The expectation changed from "we fund universities to teach and do research" to "universities should generate their own income", which isn't really possible with research, so federal funding filled the gap. This lead to the indirect fund firehose of cash, pyramid scheme labs, and so forth and so on. It sort of became a feedback loop, and now we are where we are today.
Translational research is probably part of it but I think it's part of a broader hype and fad machine tied to medicine, which has its own problems related to rent-seeking, regulatory capture, and monopolies, among other things. It's one giant behemoth of corruption fed by systemic malstructurings, like a biomedical-academic complex of problematic intertwined feedback loops.
I say this as someone whose entire career has very much been part of all of it at some level.
Good points, thanks. As I’m sure you’re aware, the indirect rates at some universities are above 90%. That is, for every dollar that directly supports the research, almost another dollar goes to the university for overhead. Much of this overhead is legitimate: facilities and equipment expenses, safety training, etc… but I suspect a decent portion of it goes to administrative bloat, just as much as the education-only part of the university has greatly increased administrative bloat over the last 30-40 years.
Another commentator made a separate point about how professors don’t always get paid a lot, but they make it up in reputation. Ego is a huge motivator for many people, especially academics in my observation. Hubris plays no small part in the hype machine surrounding too many labs.
867 comments
[ 2.1 ms ] story [ 465 ms ] threadThis is absolutely something that we should routinely be doing, though.
Pick your poison, to some extent. Better would be to not have to do it after-the-fact, but to vet better at every intermediate step, but it's hard. Just a very difficult people problem.
My concern is that with AI getting better and easier to use (e.g., in Photoshop) fraud will be extremely hard to detect.
We might need to re-think how research is done and results verified.
In theory at least research should be auditable to a lab notebook. The problem with photos and such is you can't tell if it was modified before it was pasted to the page and large datasets just can't be put in to paper. And electronic notebooks I've used tend to be even more annoying than paper (too rigidly formatted and not adaptive to workflow optimization but it's difficult to explain).
Anyway those sorts of things that establish provenance should also protect against deep learning. You may be able to create fake data, but can you deep fake data that was signed by Nikon device #XYZ with cryptographicly confirmed hashes published to a blockchain 3 years ago at the time the data was generated?
Ever so rarely, there may be an opaque, internal investigation. Mostly, it seems that academia has a desire to not make any waves, keep up appearances, and let the problem quiet down on its own.
It's like the old quote... "If you commit fraud as an RA that's your problem. If you commit fraud as the head of department that's the university's problem."
I suppose a silver lining to all the academic fraud exposés of the last few years is that more grad students and faculty now know that this is a thing, and one that many will try to cover up, so trust no one.
Another silver lining might be that fellow faculty are more likely to believe an accusation, and (if they are one of the awful people) less likely to think they can save funding/embarrassment/friend by neutralizing the witness.
(ProTip: If the success of your dishonesty-reporting approach is predicated on an internal administrator having humility and integrity, realize that those qualities are the opposite of what has advanced a lot of academic careers.)
Prediction: papers stop using pictures entirely
Accompanying to finish document will be a hash of all of these works along with their associated timestamps, originals of can be verified if necessary to prove a custodial chain of development over a plausible period of time and across multiple iterations of the development of the work - a kind of signed time-lapse slideshow of its genesis from blank page to finished product as if it had a mandatory and global "track changes" flag enabled from the very beginning - by which the entire process can be proved an original human-collaborated work and not an insta-generated AI fiction.
For example, imagine if university libraries ran storage systems based on Merkle trees with PKI signatures and researchers used those for their papers, code, data inventory (maybe not petabytes of data but the hashes of that data), etc. If there were allegations of misconduct you’d be able to see that whole history establishing that when things were changed and by whom, and someone couldn’t fudge the data without multiple compromised/complicit people in a completely different department (a senior figure can pressure a grad student in their field but they have far less leverage over staff at the library), and since you’re not sharing a database with the entire world you have a much easier time scaling with periodic cross checks (e.g. MIT and Caltech could cross-sign each other’s indexes periodically so you could have confidence that nobody had altered the inventory without storing the actual collection).
There may be newer AI methods of fraud, but it will only buy you time. As both progress, committing to record a fraud generated by technology will almost certainly be detected by a later technology.
I would guess that we're within 10 years of being able to automatically audit the majority of papers currently published. That thought must give the authors of fraudulent papers the heebee jeebies.
For the same reason, tools that try to detect AI-generated text are ultimately going to lose the arm's race.
A scapegoat is something that takes on all the sins of a lot of others who skate free. If Masliah is the only one who ever suffers, then he IS a scapegoat, but if this article serves to uncover a lot of other bad actors, then he's not. And if his example serves to warn a lot of other scientists to clean up their acts, then his suffering is a benefit.
Correction: there is a new scoundrel every month. It would be nice to expose them all instantaneously, but unfortunately that's not possible.
All of them
Nassim Taleb said that people think weathermen are terrible predictors of the future. He says meteorology is among the most accurate sources of predictions in our lives. But we can easily validate it and we see the mistakes. If we had as much first hand experience with other types of predictions we'd appreciate the accuracy of weatherman. My point is: just because you know the flaws in a system don't assume it isn't better than another.
Like many others, I hope that a greater focus on reproducibility in academic journals and conferences will help reduce the spread of scientific misconduct and inaccuracy.
[1]: https://retractionwatch.com/
In my experience no, most fraudsters are not evil people, they just follow the incentives and almost non-existent disincentives. Scientist has become just a job, you find all kinds of people there.
As far as I know no-one goes to jail, worst thing possible (and very rare) is losing the job, most likely just the reputation.
Maybe I'm too idealistic but why does following incentives with no regard for secondary consequences not evil?
I don’t actually know what people mean when they label someone as “evil”, other than “is doing/saying/thinking stuff I find very reprehensible”. Which doesn’t make sense when you insert it into the above statement: “Are people who commit this type of science fraud just humans who do stuff I find really reprehensible?” Well, I guess it sounds like they are.
It seems like people want to assign a character trait when they say “person X is evil”, but I don’t believe such a generic character trait exists (and what exactly it is supposed to mean if it existed). What’s worse, it obfuscates and prevents understanding the actual character traits and circumstances that lead to the respective behavior.
However, by removing empathy, we also reduce the possibility to understand the human motivations behind heinous acts (there always are), find solutions, build bridges, make truces, end wars. So maybe we should go lightly on the "evil" stuff, as much as possible.
The concept is emotionally laden and ill-defined, and has little relation to why the designated behaviors actually happen. It’s an incoherent concept that has no explanatory power.
Socrates got it. "I know that I know nothing" (else)
You'll see this all over risky startups. What starts as hopeful optimism only becomes fraud over time, when the consequences of not committing fraud also seem horrible. It's easy to follow the road until all your choices are horrible in different ways, and they pick the one better for the people around them, yet worse for everyone else.
Lack of foresight isn't a virtue, it's as much of a vice as knowing the consequences and ignoring them. If you lack foresight and that causes you to commit fraud, you committed fraud, plain and simple. That is evil.
This kind of fraud happens because scientists are rewarded greatly for coming up with new, publishable, interesting results. They are punished severely for failing to do that.
You could be the department's best professor in terms of teaching, but if you aren't publishing, your job is at risk at many universities.
Scientists in Academia are incentivized to publish papers. If they can take shortcuts, and get away with it, they will. That's the whole problem, that's human nature.
This is why you don't nearly as many industry scientists coming out with fraudulent papers. If Shell's scientists publish a paper, they aren't rewarded for that, if they come up with some efficient new way to refine oil they are rewarded, and they also might publish a paper if they feel like it.
A lot of companies reward employees for publications. Mine certainly does. Also an oil company may not be such a great example since they directly and covertly rewarded scientists for publishing papers undermining climate change research.
A scientist can work in industry research and NEVER publish, and still have a career where they make money, and don't worry about losing their job.
Once it became a distinguished profession the incentives changed.
"When a measure becomes a target, it ceases to be a good measure"
From a distance the practice of science in early modern and Enlightenment times might look like the disinterested pursuit of knowledge for its own sake. If you read the detailed history of the times you'll see that the reality was much more messy.
For example Kepler considered his work on the Music of the Spheres (google it) to be more important than, and the ultimate goal of, his research on the mechanics of planetary motion. Newton dabbled in alchemy, and his dispute with Leibnitz was very very bitchy with some dubious jostling for priority. And there was no end of dubious research and outright fraud going on at the time. So no, it was not a golden era of disinterested research.
See for example the wikipedia articles on Phlogiston, The Music of the Spheres, the long and hard fought battle over Epicycles etc
As a quick example, any book about electricity from the early 1900's will include quite serious sections about the positive effects of electromagnetic radiation (or "EM field therapies"), teach you about different frequencies and modulations for different illnesses and how doctors are applying them. Today these devices are peddled by scammers of the same ilk as the ones that align your shakras with the right stone on your forehead.
The fact that you can recall those reinforces the point that the value is determined by how long it is useful and remembered, not the fact that it was published.
There are of course incredible scientists that went down disappointing paths (eg Shockley, Dyson, Pauling) in terms of their research output later on, though one must remember that typically this occurs outside their original field of expertise.
If you read my comment you will see that I am asking for the references to the claims the previous author made. I simply provided my own references which werew written at the time and are representative of the times that do not corroborate the tall tale of the previous author. If you have any references to support their claim I would be interested in perusing them.
Picking the winners as examples is not good sampling.
Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted. I think that this assertion should be backed by considerable evidence, and that burden is of course mostly on you
I don't dispute that there were doctors applying electricity and/or magnetism to the body in an "un"-rigorous manner, but is there documentation that suggests that the scientists at the time had come to the conclusion that it worked?
Also notably, Whittaker's work was a 'loser'. I chose it specifically for this purpose. I had read parts of it previously because it was a 'loser' as he chose to dispute Einstein's contributions to special relativity.
I will say that
> Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted.
Is not correct as far as I believe. Instead, we're saying that there's no reason to believe any study until it has stood the test of time. The longer it remains impactful the better.
I am building from this statement:
> The scientific method works over a long period of time, but to blindly trust a peer review study that just came out, any study, is almost as much faith as religion, specially if you're not a high level researcher in the same field and have spent a good amount of time reading their methodology yourself.
Saying that textbooks from Maxwell's era had misunderstandings and bad information is not saying they are inept, it's saying that that is how it works, it always has and always will be that way. That's it, really. The fact that good science came from it is to be expected, and the fact that bad science existed is also not to be suprising.
I think you interpreted the statement about the 1900s textbooks being wrong as a slander against the entire era, which is not how I read it, and certainly not what I meant to imply by any of my comments.
Upon rereading I agree, I apologize to you and OP for my misunderstanding. However, ultimately in general I still have to disagree at least semantically with "standing the test of time". I am not really familiar with the processes in biological or social sciences, but from a physical science background, any result of interest will need to be built upon quite quickly. Either some kind of design will be reproduced to improve it or use it, or in the case of a theoretical result it will be awaiting some kind of experiment to validate it.
>> The scientific method works over a long period of time, but to blindly trust a peer review study that just came out, any study, is almost as much faith as religion, specially if you're not a high level researcher in the same field and have spent a good amount of time reading their methodology yourself.
I don't necessarily disagree with this statement (besides the 'long period of time'). Though I would also say that simply mistrusting the result has the same issue, so the only correct way forward seems to me to be to act as if it does not exist until you gain the expertise.
>Saying that textbooks from Maxwell's era had misunderstandings and bad information is not saying they are inept, it's saying that that is how it works, it always has and always will be that way. That's it, really. The fact that good science came from it is to be expected, and the fact that bad science existed is also not to be suprising.
I'm not familiar with textbooks of that era as through personal curiosity I've only read a few. I would still like to see an example of such an occurrence to understand the context under which these treatments are discussed. If these fallacious techniques were widespread enough to be popular in textbooks there must be some kind of literature supporting them?
>I think you interpreted the statement about the 1900s textbooks being wrong as a slander against the entire era, which is not how I read it, and certainly not what I meant to imply by any of my comments.
I will admit to being a bit hotheaded in the initial response, which I apologize for.
>Again, you and the original poster seem to have this understanding that scientists and engineers from the mid 1800s to early 1900s are not to be trusted.
No, once again that's not what was said. The concept being communicated is that the scientific method works over long periods of time, not short ones. Over long periods of time, such as 200 years, the work that survives peer review and remains significant today are things like Maxwell's work on electromagnetism as opposed to Dr. Franz Mesmer's work on animal magnetism.
You are taking little bits and pieces of what people are saying, misconstruing them and reinterpreting them, and then forming an argument that is not a genuine representation of the original comment.
I will admit to some mistakes in comprehension and a poor literal quoting, though I will also maintain that I captured the majority of the essence of what was written in the quote. In the case of "any book about electricity..." vs "any engineering book" the only books that would be relevant to the discussion should be engineering OR science books relating to electricity.
>No, once again that's not what was said. The concept being communicated is that the scientific method works over long periods of time, not short ones. Over long periods of time, such as 200 years, the work that survives peer review and remains significant today are things like Maxwell's work on electromagnetism as opposed to Dr. Franz Mesmer's work on animal magnetism.
With respect to "time" I will refer to another comment I made below in response to the previous OP.
I think it's interesting that you use Mesmer as an example because his work failed to gain the acceptance of the scientific societies of the time, and was in the late 1700s, significantly earlier than the proposed mid 1800s to early 1900s
This might seem crazy to hear now, but when Maxwell first published A Dynamical Theory of the Electromagnetic Field in 1865, no one cared, it received very little attention at the time.
It was decades later in 1888 with the work of Hertz that Maxwell's equations started to gain significance within the scientific community.
I think you will also find that the publications of the 1800-1900s are quite well preserved.
You posted this less than 15 minutes after my comment, friend.
> I assume that the reader is familiar with the idea of extra-sensory perception, and the meaning of the four items of it, viz. telepathy, clairvoyance, precognition and psycho-kinesis. These disturbing phenomena seem to deny all our usual scientific ideas. How we should like to discredit them! Unfortunately the statistical evidence, at least for telepathy, is overwhelming. It is very difficult to rearrange one's ideas so as to fit these new facts in. Once one has accepted them it does not seem a very big step to believe in ghosts and bogies. The idea that our bodies move simply according to the known laws of physics, together with some others not yet discovered but somewhat similar, would be one of the first to go.
Anyone on this site who doesn't know what this is from should feel a bit of shame in the current era of hype around machine intelligence, so I'll leave it as an exercise to the reader if you aren't already familiar with this paper.
Unfortunately he is out of his area of expertise in physics and human biology/neuroscience (? not sure where telepathy would be if it was to be rigorously studied). This is akin to Freeman Dyson on global warming.
That scientists can have strange ideas is something nobody can dispute. That those strange ideas enter into scientific legitimacy is another story entirely.
Science even at it's most sincere should always be approached with thoughtful skepticism. The phrase that I hear touted often these days "trust the science", is in essence not how science should be thought of.
1. outside their areas of expertise
2. not validated by independent scientific research
I completely agree that science should be approached with thoughtful skepticism, and I agree that 'trust the science' might not necessarily be the best semantics to use. However, it is not clear that skepticism by all parties should be considered with equal weight. Most of the times, people should "trust the science" because they are not equipped to be skeptical.
The electro-therapeutics chapter starts on page 658
Do you have examples of usage as a treatment? I can only think of rTMS (whose effectiveness is contentious).
What is maybe the most applicable that is widely accepted is electric therapy for people recovering from ACL surgeries.
ME? the coil generates an M which induces localized E in the body as shown by localized currents? (which produce some more M, but only just enough)
This is warranted because they are pretty different - light has a frequency distribution, diffracts, etc, and can be focused to propagate energy over distances large compared to its source, whereas quasistatic fields (by definition) have no frequency distribution, and die as 1/r^2 or faster
That's not a scientific assertion though. It's just what many people would say many/most scientists believe
What is most incredible to me is even knowing and believing the above, I fall prey to this all the time.
I guess it also reinforces the supreme importance of reproducibility. Seems like no research result should be taken seriously until at least one other scientist or group of scientists are able to reproduce the result.
And if the work isn't sufficiently defined to the point of being reproducible, it should be considered a garbage study.
If you have independent measurements you cannot rule out bias from prior results. Look at the error bars here on published values of the electron charge and tell me that methodology or reproducibility shored up the result. https://hsm.stackexchange.com/questions/264/timeline-of-meas...
OP suggest was to always observe, no matter the person, which is what TFA is doing.
Feel free to expand on what you think is the problem and solution if you feel everyone is off-target.
But the OP's suggestion was that to fix this specific problem of faked observations, you should separate interpretation and observation. I don't see how that fixes this problem at all. And in my view the first step in solving the problem is to come up with some sense of how serious the problem is: meaning, rather than dwelling on each terrible isolated case and panicking, try to determine what the actual prevalance is and what the overall impacts are. With that information you can make resource-allocation decisions in how to address it. The HN response is much too emotional for anything useful to come of it (except for more anger and confirmation bias.)
Speaking of Orwell, I don't think science comes into it. Rather, when people stop believing in democracy, things will degenerate into authoritarianism. It's generally pretty hard to use science the methodology to implement an authoritarian government as the scientific method by definition will follow the evidence, not the will of a dictator.
However, something that looks like science but isn't could be used, especially if the public doesn't understand science and thus can't spot things that claim to be science but don't actually follow the scientific method.
Methodologies are inanimate - I may trust that a methodology is fine, but once Humans become involved I do not trust.
> Instead, they genuinely believe there is some alternative to finding the truth
There are several alternate means, the field of philosophy (that birthed science) has been working on such problems for ages, and has all sorts of utility, just sitting there waiting to be used by Humanity.
> and now simply believe the same old superstitions and bunk that people have prior to the scientific revolution.
Not possible for you to know, unless there are indeed forms of supernatural (beyond current scientific knowledge) forms of perception.
> Rather, when people stop believing in democracy, things will degenerate into authoritarianism.
Once again, not possible for you to know.
> It's generally pretty hard to use science the methodology to implement an authoritarian government
COVID demonstrated that to be incorrect.
> as the scientific method by definition will follow the evidence, not the will of a dictator.
Incorrect. Something defined to be true necessarily being true only works in metaphysics, such as linguistics.
And again, the scientific method is inanimate.
> However, something that looks like science but isn't could be used, especially if the public doesn't understand science and thus can't spot things that claim to be science but don't actually follow the scientific method.
On a scale of 1 to 10, how comprehensively and accurately do you believe you understand science?
Great critical thinkers become lawyers, post modernist intellectuals, and other parts of the "talking" class of intellectuals. Unfortunately, it's far easier to talk shit than it is to build things. We've massively over-valued critical thinking over constructive thinking.
Most people want to dunk on science. Few people want to submit their own papers to conferences. Many people act like submitting papers is impossible for non-Ph.D's. We have a lack of constructive oriented thinking.
I agree. But academia reinforces this perception. I feel like PhD's only give serious consideration to the utterances of other PhD's. The rest of the public consists of the unwashed masses, and at best gets the smiling-nod treatment from teh PhD.
PS I (a non-PhD) managed to publish a paper during the pandemic (doi: 10.3389/fphar.2022.909945 ). One of the biggest barriers was the item you mentioned quoted above, and the bogeyman of "epistemic trespass" in general, as operating in my own psychology. I've since become noisy in advocating for the #DeSci movement.
I wonder if it's possible to get people to wonder why the one discipline that has the tools to deal with all of these epistemic, logical, etc issues isn't taught in school. You'd think it would be something that people would naturally wonder about, but maybe our fundamentalist (and false) focus on science as the one and only source of knowledge has damaged our ability to wonder independently.
This does not seem necessary for my ask above.
There may be many approaches, some impossible/invalid, but perhaps not all.
It is not fine to put trust into scientists in general just because they walk around in a lab coat with a PhD label on its front.
For most of the people who end up in these scandals, this is just the day job that their various choices and random chance led up to. they're just ordinary humans responding to ordinary incentives in light of whatever consequences and risks they may or may not have considered.
Other careers, like teaching, medicine, and engineering have similar problems.
Indeed, I believe many academic careers were kicked off in this manner. Where it all goes wrong is when other more diligent researchers fail to reproduce said fraudulent research - this is what brought down famous fraudster Jan Hendrik Schön in the field of plastic-based organic electronics, which involved something like 9 papers in Science and Nature. There are good books and documentaries on that one. This will only be getting worse with AI data generation, as most of those frauds were detected by banal data replication, obvious cuts and pastes, etc.
However, when you add a big financial driver, things really go off the rails. A new pharmaceutical brings investors sniffing for a big payout, and cooking data to make the patentable 'discovery' look better than it is is a strong incentive to commit egregious fraud. Bug-eyed greed makes people do foolish things.
This is why I think the most effective way is to empower good actors. Ensure open debate, limit the power of individuals, and prevent over concentration of power in a small group. These efforts are harder to implement than you think because they run against our desire to have scientific superstars and celebrities, but I think they will go a long way towards building a healthy community.
Anytime you see champagne bottles up on a professor's top shelf with little tags for Nature publications (or something like that), then you know they are a glory hound.
When you see beer bottles in the trash, then you know they're in it for more than themselves.
Jobs at top institutions are worth much more than their nominal salary, as evidenced by how much those people could be making in the private sector. (They are compensated mostly in freedom and intellectual stimulation.) Unambiguously faking data, which is the sort of thing a bad actor might do to get a top job, should be considered at least as bad a moral transgression as stealing hundreds of thousands or perhaps a few million dollars.
(What is the downside? I have never once heard a researcher express feeling threatened or wary of being falsely/unjustly accused of fraud.)
Are you going to go to jail for fudging some numbers on your paper, not likely.
It's bothering you a rational amount, actually. These people have done serious damage to lots of lives and humanity in general. Society as a whole has at least as much interest in punishing them as it does for financial fraudsters. They should burn.
When the experiments are significant, double blind is not enough. You need external auditors when conducting experiments. Preferably separate team making experiments from those who design them.
Ex: Finding when the exact same image appears in multiple publications, but with different captions/conclusions.
The evidence in this case came from one individual willing to volunteer hundreds of hours producing a side by side of all the reports. But clearly that doesn't scale.
How would this work? AI can't even detect AI generated content reliably.
- Parse all papers you want to audit
- Extract images (non AI)
- Diff images (non AI)
- Pull captions / related text near each image (LLM)
- For each image > 99% similarity, use LLM to classify if conclusions are different (i.e. highly_similar, similar, highly_dissimilar).
Then aggregate the results. It wouldn't prove fraud, but could definitely highlight areas for review. i.e. "This chart was used in 5 different papers with dissimilar conclusions"
It's fraud all the way down, where even the fraud detectors are fraudulent. Similar story to the anti-malware industry, where software bugs in security software like CrowdStrike, Sophos, or Norton cause more damage than the threats they prevent against.
[1] https://www.reddit.com/r/ChatGPT/comments/11ha4qo/gptzero_an...
Or, in the other hand, now you don't have to manipulate images, you can just generate the ones you need.
I mean, get this: an extremely gifted Mathematician living on a measly salary in Russia had had his millenium prize almost stolen by a Harvard professor. What more evidence do you need?
My advice to anyone going into Hollywood would be to stay away from drugs; my advice to anyone going into academia is to treat every interaction as if you've just sat at a poker table in Las Vegas.
Our experience of things has a lot of bias toward what we want to hear. Generalization plays into sterotypes and ideology.
edit: and it's not things I've heard, instead it is direct experiences, i.e. people stole my work, and things like that. As a graduate student to watch professors come to you with problem X, take what you've said (an actual solution) and publish a paper without attribute, that sort of thing; to report it and have nothing be done about it, et cetera, and on it goes, it's just instance after instance of such behavior, or the million ways in which they are careful to trick you into working on their problems without receiving attribute. One such trick for example, that again happened to me, is that after a conference talk I got into an e-mail discussion where I explained my approach; I was told that "they already have these results" (the trick here was to divulge less in the talk than what was currently known in order to be able to avoid "significant progress by another person" in the case another person does share new progress that they have already established, and hence not having to share attribution.) It turned out that our discussion was enough for them to go from n=3,4 to a general formula involving primes, because I pointed out a certain property they had not noticed. This is just a single example of the sorts of tricks, aside from total fraud, that happen, and one of the milder incidents I had happen to me.
If you are unable to "extrapolate to all of academia" then I suggest you be more selective in your statements.
I blame this behavior entirely on "publish or perish". The demands for novel, thoughtful and statistically-significant findings is tremendous in academe, and this is the result: cheating.
I left professional academia because I resented the grind, and the push to publish ANYTHING (even reframing and recombining the same data umpteen times in different publications) in an effort to earn grants or attain tenure.
The academia system is broken, and it cannot be repaired with minor edits, in my opinion. This is a tear out and do over scenario for the academic culture, I'm afraid.
We need to seriously rethink our approaching to stewarding these institutions and ideas, public trust is rightfully plummeting.
Want to publish in a peer reviewed paper? Well then your institution or you should take out a bond or insurance policy that guarantees your work is accurate. The insurance amount would fluctuate based on how big of impact this study could have. Is it a drug that will be consumed by millions? Big insurance policy. Is it a behavioral study without much risk... small insurance policy.
Is a a person in an institution found caught committing fraud, well now then all papers from that institution now have higher premiums.
Did you sign off on a peer reviewed paper that was fraud? Well now your premiums are going up also.
Insurance costs too high to publish? Well then keep doing research until the underwriters are satisfied that your work isn't fraud and adjust the premiums down.
It adds a direct near-term economic incentive to publish honestly and punishes those that abuse the system.
The insurance model does not really work when the cost of evaluating the risks far outweighs the expected risks.
The problem could be, that it may become impossible to publish certain kinds of papers that are very well supported and valuable because no institution can afford the insurance.
What type of research would that be? Just publish it online without insurance and everyone will treat it as it unverified and uninsured... separate from other research that is.
Once the risk of the publishing research has gone down (i.e. reputable peers approve, or the findings were replicated), the cost of the insurance goes down also.
if something is so costly to insure, there would be a reason and thus the system works.
The "organic" way this would happen is if there was a shift so that journals with insured research are far more valuable than uninsured research. Or perhaps if companies started suing researchers for negligence and fraud and recuperate costs if they used research that was later proved to be fraud.
In the literary world, anyone can publish a book, but a book from o'reilly caries with it a different level of authority and diligence then a self published book or blog post.
So the shift would have to be that your career can't advance without publishing a bonded and insured paper.
Science is largely about doing novel things and often being the first person in the world to try something. In order to understand the risks, you have to understand the actual research, as well as the personalities and personal lives of the people doing it.
Then there is the question of perverse incentives. Research fraud is not a random event but an intentional action by the people who take the insurance. If they manage to convince you to underwrite their research, they know that the consequences of getting caught will be less severe than without the insurance, making fraud more likely. Normally intentional fraud would not be covered by the policy, but here covering it would be the explicit purpose of the insurance.
The research might be novel, but the procedures for research and publication are very similar. So insurance companies would just make sure that you followed a protocol which minimizes their risk.
perverse incentives are taken into account by insurance. Insuring someone is always a adversarial back and forth to determine if they are being truthful or not. Which is why Life insurance companies require a physical. They don't just have you self report and then accept it as fact.
Industry professionals like lawyers and doctors carry malpractice insurance. A lawyer can still commit fraud. Insurance isn't a black and white thing. It is a sliding scale that ties risk to a monetary value.
Its not rocket science. Just actuarial science. ;)
This is wrong.
Some time ago, I completed the checklists for publishing a paper in a somewhat prestigious multidisciplinary journal. Large parts of the lists were about complying with various best practices and formal requirements in different fields. I often didn't even understand the questions outside my field. And the questions nominally within my field were often category errors. They assumed a mode of doing research that was far from universal. Overall, the process was more frustrating than (let's say) applying for a US visa.
Yes there is novel research that has never been done before? So what? That doesn't change if you can get insurance or not. Thats a failed argument from the beginning.
Anyways you don't seem to be having a discussion in earnest and instead you seem to be intentionally disregarding large pieces of the above arguments and trying to shoehorn in your idea that if there is unique research being done that it means that it is impossible to tell the risk of anything. Kinda silly.
Additionally, even if we assume that the insurance model is a good idea, it should be tied to individual researchers, not universities. The entire model of university research is based on loose networks of independent professionals nominally employed by various organizations. Universities don't do research, they don't own or control the projects, and they don't have the expertise to evaluate research. They are just teaching / administrative organizations that provide services in exchange for grant overheads.
Absolutely not. Underwriters are smart. They use other variables and methods for determining risk. They don't need to directly recreate and peer review the research themselves.
1) revoke all of their academic accreditations and degrees
2) put them on a public “do not publish” list permanently banning them from being named on any paper in a journal
But if a health care professional does the same thing, and does something negligent, then they are usually liable. They are professionals and are held to a different standard. Similarly, that's why lawyers keep writing: this is not legal advice and you are not my client.
Perhaps a professional in science should have higher standards. Obviously they shouldn't be sued for being wrong - that would destroy science, disregard the scientific method's means to address inaccuracy, and go against science's nature as the means to develop new knowledge. But intentionally deceiving people perhaps should be illegal and/or create liability: When you publish something, people depend on its fundamental honesty and will act on it.
To put the scale of this problem in perspective, the ORI was set up in the 1970s after Congress became concerned at widespread reports of scientific fraud. It clearly didn't work, but hangs around regardless.
It's ultimately a culture problem. Until academics have the same level of respect as ordinary corporate employees, you're going to get judges and juries who let them off scott free.
I wrote a blog post on how to make this easier, including a new criminal statute specifically tailored for scientific fraud. https://news.ycombinator.com/item?id=41672599
The guy said, "That depends on whether you consider reputation 'nothing.' "
I guess what that shows is, you can always negotiate and compromise over money, but reputation is more of a binary. An academic can fake some work, and as long as he's never called on it, his reputation is set.
So yeah, a little more fear of having one's reputation ruined would go a long way towards fixing science.
But that one-dimensional view is boring. Life is more than politics.
And I'm not sure what example could illustrate the problem with the lopsided valuation of economic capital and the general devaluation of symbolic capital (as compared to pre-1980s, we have since undergone a social revolution of considerable dimensions, which is also why there isn't an easy fix) better than this one.
Humans are, by our biology, social creatures. Modern humanity more than ever before. If you're not considering the social effects, then IMO you're not addressing anything of value.
Bang on.
Not many people in the academic/technical people realize this, often for their entire lives. In their naive worldview, they cannot even imagine that people can stoop that low.
(embarrassingly and shamefully I used to be one of those naive people)
A caveat that "reputation", like competence, is more variagated and localized than is often appreciated. As with someone who is highly competent and well regarded in their own subfield, while simultaneously rather nutter about some nearby subfield where they don't actually work.
One can have a reputation like "good, but beware they have a thing for <mechanism X>". Or "ignore their results using <technique> - they see what they want to see". Subtext that gets passed among professors chatting at conferences, and to some extent to their students, but otherwise isn't highly accessible.
When people speak of creating research AI's using just papers... that's missing seemingly important backchannels. And corresponding with authors. Attempting research AI as developing-world professionally-isolated professor.
Defund universities. No more student loans, make them have to earn their place in the market or we will continue to suffer under the manipulated system that is actually killing students.
This... it's no longer about value its about optics... Problem exists in most industries now. The pendulum needs to swing back the other way before it's too late to stop the decay...
Every year that feeling becomes more certain. Glad I quit the track in grad school.
I feel terribly for all the incredibly smart and hard working academics that remain honest and try to make it work. They do what they love, otherwise they wouldn't do such intensive work with so much sacrifice.
It is really disheartening too because academia only turns on the "honesty filter" when it comes to minor grad students that pissed off the wrong people. But you can do all this fraud constantly and become president of harvard if you know the right politics.
Dishonest lot. I hope karma is real so they get what is coming to them for taking advantage of people that just love to increase humanity's knowledge.
You're being downvoted because you're correct—HN is an eco chamber for zealous regurgitation of opinions of the academy and media—institutions that have decayed. It's been happening slowly for awhile, but now things are starting to come apart at the seams.
"We know academia is bad. But this is the best we have and it is hard to improve"
when that is false on two counts.
1. If you had said the same thing before 2016 or covid, people would not agree that academia is rife with fraud or worthy of skepticism. 2. The same people saying the system dismiss how can it be improved are the same ones that would suffer from disruption as you say. They have the power to dismiss these arguments to begin with.
It's profoundly insulting and grotesque—on par with excusing the inexcusable.
Accepting this degree of mediocrity is as repulsive as tolerating the most heinous acts imaginable. I've confronted people directly with this, to their face, because to me, it's inconceivable how anyone can be okay with such a vile acceptance of the status quo.
If society was even slightly capable of rational action . . . (legally, I cannot complete this sentence).
I can do it too. Person named SpaceManNabs made a bad post. There for all posts by SpaceManNabs, and probably all posts on HackerNews are bad. A dishonest lot.
I specifically mention that the majority of scientists are not dishonest. The majority of scientists are not running academia. The majority of scientists are suffering from this system, to differing degrees.
If I were as rude as you, I'd extrapolate on reading ability, especially since it is not just one fraud case.
Regardless, even if I was wrong on that, all my other criticisms of academia still stand, like exploitation of the phd students. I really hope the grad student unions get what they want.
I appreciate your response though. Makes me feel confident that it is just salty people on HN that hate truth, because otherwise, why would you mischaracterize what I said?
It all comes down to lab notebooks and data policies. If there is no system for archiving detailed records of experimental work, if data is recorded with pencils so it can later be erased and changed, if the PI isn't in the habit of regularly auditing the world of grad students and postdocs with an eye on rigor and reproduciblity, then you should turn around and walk out the door immediately.
As to why this situation has arisen, I think the corporatization of American academics is at fault. If a biomedical researcher can float a false claim for a few years, they can spin their research off to a startup and then sell that startup to a big pharmaceutical conglomerate. If it fails to pan out in further clinical trials, well, that's life. Cooking the data to make it look attractive to an investor - in the almost completely unregulated academic environment - is a game that many bright-eyed eager beavers are currently playing.
As supporting evidence, look at mathematical and astronomical research, the most fraud-free areas of academics. There's no money to be made in studying things like galactic collisions or exoplanets, the data is all in the public domain (eventually), and with mathematics, you can't really cook up fraudulent proofs that will stand the test of time.
Is there evidence of the fraud levels in those fields?
So we're systemically safeguarding the quality of astronomy research, by setting up a gradient (at MIT: restaurant catering for business talks, pizza for CS, stale cookies for astronomy) to draw off some flavors of participants and thus concentrate others?
You are talking about a part of the academy that relative to medicine, so few people do.
Show up to a bank looking like someone who knows math, and they'll cut you a huge check. Is that not fraud?
Sure money could be a factor, but the desire for prestige can motivate people just as easily.
On one hand, it seems the push for immediate real world relevance is a good thing. We fund research in order that society will benefit, correct? On the other hand, since publications and ultimately funding decisions are based on demonstrating real world relevance, it’s little surprise scientists are now highly incentivized to hype their research, p-hack their results, or in rare cases, commit outright fraud in an attempt to demonstrate this relevance.
Doing research that has immediate translational benefits is a tall order. As a scientist you might accomplish this feat a few times in your career if you’re lucky. The rest of the corpus of your work should consist of the careful, mundane research the actual translational research will be based upon. Unfortunately it’s hard to get that foundational, basic, research published and funded nowadays, hence the messed-up incentives.
Translational research is probably part of it but I think it's part of a broader hype and fad machine tied to medicine, which has its own problems related to rent-seeking, regulatory capture, and monopolies, among other things. It's one giant behemoth of corruption fed by systemic malstructurings, like a biomedical-academic complex of problematic intertwined feedback loops.
I say this as someone whose entire career has very much been part of all of it at some level.
Another commentator made a separate point about how professors don’t always get paid a lot, but they make it up in reputation. Ego is a huge motivator for many people, especially academics in my observation. Hubris plays no small part in the hype machine surrounding too many labs.