Everyone is Cheating (Even the Professors)

This video is sponsored by Private Internet Access. Stick around to the end

Internet Access. Stick around to the end to find out more. A student from Nor Eastern University demanded her tuition back after she discovered that her professor was using Chat GPT to teach the class. It wasn't fair, she said,

the class. It wasn't fair, she said, because this professor had banned students from using AI while he was using it himself. At another university, a student got some good news. She had

gotten an A on a paper. So then she went to go and look at the comments that the professor had left on it, and she found that the professor had copy and pasted an exchange with chat GPT. Not only was the professor using AI to grade the

paper, but they were sloppy about it, too. More and more students are using AI

too. More and more students are using AI for all of their work. And as one writer at New York Magazine put it, "It seems like everyone is cheating their way through college." The way people tell

through college." The way people tell this story, AI is ruining education. And

on top of that, Ginz is too lazy to do any work. But cheating at universities

any work. But cheating at universities is nothing new. It's a long-running issue, and it has to do with the incentives of education in the United States. And it even involves the

States. And it even involves the widespread issue of academic fraud from professors. We'll explore all of that

professors. We'll explore all of that today, starting with this latest problem AI.

When Open AI was founded, it had a mission. It wanted to ensure that all of

mission. It wanted to ensure that all of humanity benefited from artificial general intelligence. If you read a lot

general intelligence. If you read a lot of the early writing that was coming out of OpenAI, it sounds rather utopian.

They were saying that they were a nonprofit. They would be focused on just

nonprofit. They would be focused on just doing research and would even be able to give away that research so that it would benefit the world. All of humanity should benefit from this thing they wanted to create. true AI. Open AAI

might be a nonprofit, but within OpenAI, there's a for-profit company, and it's this for-profit company that tries to bring products to market. And OpenAI is now worth something like $300 billion.

And that's because people really want what they're selling. Every major

artificial intelligence company now heavily markets themselves towards students, and it's definitely working.

Nearly half of students and teachers at the K through 12 level say that they use AI, and nearly every college student uses AI in some capacity. The result is that we have issues of plagiarism,

cheating, and fraud. AI can write papers quickly. It can make it easy to pass a

quickly. It can make it easy to pass a class, and it frees up a student's time so that they can do stuff that they'd rather be doing. Whether that's working a job or thinking about their career prospects or just, you know, going out

and drinking or something. There are now apps, websites, other services, all dedicated to helping students use AI to cheat and pass their classes. Professors

who want to avoid cheating have to find workarounds. They'll sneak in extra

workarounds. They'll sneak in extra words into assignments that are designed to mislead AI. Or they'll demand that students take exams with blue books, like literally writing a test with, you know, by hand with a pencil or a pen.

This might solve part of the problem, but when it comes to things like take-home assignments or extended essays or even short reading responses, well, the overwhelming odds are that students

are using AI to do basically all of that work. The consequences of this are

work. The consequences of this are pretty dire. One professor at Cal State,

pretty dire. One professor at Cal State, Chico, said that massive numbers of students are going to graduate from university and they're going to enter the work and they will be essentially illiterate. I think the concern is that

illiterate. I think the concern is that whatever skills we think you're supposed to develop when you go to college, when you go to university, aren't being developed at all. That's because more students are finding out they don't need

to read or write or have any original thoughts in order to get their degrees.

Some universities like Ohio State have decided that the solution is to just embrace AI completely. Ohio State

actually says that every student will be encouraged to use AI in every class from now on, but most universities simply don't know what to do. The philosopher

Megan Fritz, who teaches at the University of Arkansas at Little Rock, wrote an article about this. Fritz sat

on several university committees that were trying to come up with AI policies that could help, you know, solve some of these problems at her school. And they

got nowhere. They tried to develop policies for departments, but they couldn't keep up with the progress from various AI companies. the technology was just moving too quickly. And in an essay for the point, she wrote that the

biggest difficulty was determining what the goal of an education was supposed to be. And if you can't answer that

be. And if you can't answer that question, then you probably can't answer any of the questions about AI and cheating.

So in his book, The Case against Education, the economist Brian Kaplan makes the case that Americans spend too much time and too much money on education in general. I would argue

against the main thrust of Kaplan's book, I think, but he does make an early point that's really important for this discussion. Much of modern education is

discussion. Much of modern education is really not about passing on skills or learning things. It's simply about

learning things. It's simply about getting the degree at the end of the day. And that's because a degree signals

day. And that's because a degree signals certain facts about you. And Kaplan says that those include intelligence, conscientiousness, and conformity. Many

employers will require a college degree for a high-paying job, but they're not necessarily looking for specific skills that you acquired in the course of getting that degree. An art history major might end up working on Wall Street. After I got a PhD in philosophy,

Street. After I got a PhD in philosophy, I went into tech. And it's not like my PhD in philosophy was directly setting me up to go work in tech. Many people

get degrees in things like communications or psychology. And they

get jobs that do not require you to know anything about theories of communication or facts about the human mind. By the

way, my new puppy Apollo is joining us for this video because he really likes to be held. If the skills you acquire in the course of getting a college degree don't actually matter, then why do employers require one? It's because a

college degree does usually signal some facts about the job seeker. It just has very little to do with the content of your education. It basically says that

your education. It basically says that you were able to survive for 4 years at an institution that has kind of arcane and arbitrary rules and that you were good enough at following those rules and

smart enough to navigate them that you could get a degree at the end. And guess

what? Working at a large corporation is going to be an experience of surviving an arcane and obscure institution with weird rules. Kaplan's theory doesn't

weird rules. Kaplan's theory doesn't really account for specialist degrees like engineering, but we can set those aside because those are a minority of degrees. I think deep down actually many

degrees. I think deep down actually many students realize that this is true. They

look at many of the classes they have to take in college and they think why does any of this matter? All I'm looking to do is to just get a better job at the end of this. If you're going to college to increase your career prospects or to

climb the social ladder, well, you know, philosophy 101 isn't necessarily going to help you do that. And if you're a student who understands this, then using AI to cheat isn't just reasonable. It

might be the rational thing to do. At

least it is within your self-interest.

You're not there to wax philosophical about continism versus utilitarianism or to go and write an essay about Charles Dickens. I wish more people wanted to do

Dickens. I wish more people wanted to do those things, but it's not why most people go to school. Kaplan really is just thinking about the incentives of getting a college degree. And I think if there's like a big thesis for this video, it's that incentives matter. To

make Cap's point a little bit clearer, you could look at the phenomenon of diploma mills. These are institutions

diploma mills. These are institutions that set themselves up like colleges and universities. They are often for-profit

universities. They are often for-profit and unacredited. And in exchange for

and unacredited. And in exchange for some money and very little work, you can get a college degree. It's even a way to get a PhD in some cases. Some schools

will simply give you the degree if you pay them enough money. That's the most egregious example. But other schools

egregious example. But other schools will simply just make it very easy for you to get the degree. They'll count

work experience as credit, for instance.

Um, and you'll be able to get out of most of your coursework. So, you just take a couple of classes, you write 120page paper, and bam, you have a master's degree from online university.

And if all you need on your resume is just a degree from some school so that some AI filter at some company will let you through so that a human being will

actually just read your resume. Well,

then it kind of makes sense to go and get a degree. Even though these things are, you know, scam, not having a degree is a problem if you're looking for a job. A fake degree solves that problem.

job. A fake degree solves that problem.

and you don't have to learn anything.

Perhaps the tragedy of what is happening at many universities where students could just use AI to cheat on everything is that kind of all education is eventually just going to be fake and

made up because nobody actually has to learn anything to get a degree.

When I read Kaplan's book, I immediately saw a connection to another issue in academia, the widespread presence of academic fraud. This is research fraud

academic fraud. This is research fraud that's committed by professors.

Recently, there was the case of a Harvard Business School professor, Franchesco Gino. Her tenure was revoked,

Franchesco Gino. Her tenure was revoked, which almost never happens, and she was fired from Harvard after an investigation found significant evidence of fraud in her research. It seems she had used fake data in multiple studies.

And some of the papers where she faked this data was research on authenticity and honesty. It's very rare for tenure

and honesty. It's very rare for tenure academics to ever be fired. This, if you can get tenure, it is basically a guarantee of a job for life. And that

should show you just how serious these infractions were. The investigation into

infractions were. The investigation into Gino's work started several years ago, and it actually began with some science bloggers. That's where a lot of academic

bloggers. That's where a lot of academic fraud seems to be discovered. These

bloggers found significant issues in at least four of Gino's published papers.

Gino tried to say that she was a victim of defamation and gender discrimination and that her privacy had been invaded.

Then she sued Harvard. She sued the dean of the Harvard Business School, her boss, and she sued those bloggers. But

that case was thrown out by a judge. And

now Harvard Business School has determined that she really did fake data and now she's lost her job. But Gino is not alone in committing academic fraud.

Throughout 2024, there were many high-profile academics who were found guilty of plagiarism or as they sometimes put it, inadequate citation,

including a former professor of Harvard.

And just recently, a PhD student at MIT in the economics program had a paper pulled and I believe is no longer a student at MIT after the university investigated it. And they went on to say

investigated it. And they went on to say they had no confidence in the veracity of the research. Because of privacy laws and university policies, MIT really can't say much beyond that. They don't

even use the student's name in that announcement. Though they give the title

announcement. Though they give the title of the paper, and if you were to look for the title, you would easily find the author's name. This might seem fairly

author's name. This might seem fairly minor because students cheat on papers all the time. However, this paper was scheduled to be published in a good economics journal, and it had already been posted as a preprint on the website

archive. That's a website where

archive. That's a website where scientists post copies of their research so that it's more freely accessible. In

this case, the problem was that the research was fraudulent. If the paper had been legitimate, it would have made this student's career because it had a really provocative conclusion. It seemed

to show that using AI actually boosted a firm's research productivity. So

incorporating AI would actually make employees more productive. This is a result that many people who are big fans of AI would love to see. It would seem to be a great response that like skeptics of generative AI. People were

naturally interested in this paper and it was getting a lot of buzz from the moment it was posted on archive. And if

you're a graduate student, that's the sort of paper that you would want to publish as you go into the job market.

It's the sort of paper that would identify you as a young and brilliant scholar who was going to produce great research and you would definitely get a job. You could be set for life. based on

job. You could be set for life. based on

a paper like that. And sometimes this academic fraud has real world consequences. One of the most famous

consequences. One of the most famous academic frauds is Brian Wansync.

Wansync was a nutrition scientist at Cornell and he's actually responsible for some studies that entered the public consciousness. You might believe some of

consciousness. You might believe some of the things that was initially discovered that's discovered in quotes by Wansync and his team. Have you ever heard that shopping on an empty stomach is a bad idea because you'll buy more food or

maybe you'll buy more junk food? that

was originally from a wansing study and I believe that study has been discredited along with much of his other research. He was also responsible for

research. He was also responsible for that idea that the size of your plate determined how much you wanted to eat at a meal. In 2017 it was discovered that

a meal. In 2017 it was discovered that Wansync had in some cases fabricated data and in other cases he had done misleading analyses of the data to try and support his claim. The the simplest

way to describe that though is that he committed fraud. He was actually found

committed fraud. He was actually found out because he wrote a blog post about his own methodology. And he admitted that sometimes he would have graduate students go over experiments that seem to be initial failures and go looking

for interesting results in the data. And

this is a classic sort of research misconduct. And Lancync like Gino was

misconduct. And Lancync like Gino was only discovered because some science bloggers got interested in the issue.

They started looking into it and then once they found something interesting they kept digging. When Cornell reviewed his research, they found many problems, including misreporting research data, problematic statistical techniques,

failure to properly document and preserve research results, and inappropriate authorship. When we're

inappropriate authorship. When we're talking about student misconduct at universities, like using AI to cheat in all their classes, I think it's also important to look at professors and ask questions about why do so many

professors commit research fraud? Or you

could ask the more general question, just why do they lie?

We like to think of scientists as noble individuals in pursuit of truth. How are

they tempted to abandon this pursuit of truth and then go and commit fraud? One

of the best writers on this subject is Liam Kofi Bright. He's a philosopher at the London School of Economics and he studies the social epistemology of science. In other words, he studies how

science. In other words, he studies how the institutions and group dynamics of science affect scientific progress. And

in a paper, why do scientists lie? Brite

points out that if you have a PhD in physics or some other science, you could actually go and get rich by leaving academia and entering industry. If

you're looking to get rich as a scientist, you're better off just leaving academia. And that would suggest

leaving academia. And that would suggest that most professors don't commit fraud because they're greedy. If they were really greedy and just wanted more money, they would just quit their university jobs. Wright instead defends

university jobs. Wright instead defends an idea that's originally due to Robert Mertin. Mertin theorized in the 1960s

Mertin. Mertin theorized in the 1960s that scientists are what we could call credit seekers. Another term that Bright

credit seekers. Another term that Bright and Martin use is honor. They seek

status in the scientific community. The

way we have arranged our institutions and you achieve status by producing research and we typically want novelty.

So we want the results to be new and volume of output. Most academic

researchers go through something called a tenure review at some point in their career. This is where academics evaluate

career. This is where academics evaluate your work and they decide whether or not you should be granted tenure. And if you get tenure at a school, you are basically guaranteed a job for life.

You're not going to become fabulously wealthy this way, but it essentially guarantees an upper middle class existence. And if you fail your tenure

existence. And if you fail your tenure review, typically you're fired. And to

pass your tenure review, you prove that you're a good researcher and and two of the major metrics we use to judge this are the number of articles or books that you've published and how often you're cited by other people. So essentially,

we're asking, do you produce a lot of research? And is that research

research? And is that research interesting enough that people keep talking about it when they do research?

Turns out that both of these metrics can be gamed. There are journals that you

be gamed. There are journals that you can pay to publish your academic papers.

It's fairly easy to get caught doing this, but some academics still do it.

Another way, though, is to commit fraud.

And most academics who commit research fraud probably don't think they're doing anything that bad. And they're not going to do things that are as egregious as some of the people that we've talked about today, like where they just fake data and make things up. So instead,

they're fudging statistics. They

slightly misdescribe the evidence. They

might even believe that they're doing nothing wrong. When you're properly

nothing wrong. When you're properly incentivized, you can deceive yourself.

If you want to then get cited more, well, you can pay for that, too. You can

still get caught. So, again, if you want to avoid scandal and boost your citations, you just need to publish new and interesting ideas. Take a look at how Brite puts it. Scientists want and need credit for new results to ensure

their results are new, i.e. novel enough

to have a plausible claim to establishing priority. Scientists have

establishing priority. Scientists have to do their work sufficiently quickly that nobody else beats them to the punch. So because these scientists have

punch. So because these scientists have to work fast, they are incentivized to either be sloppy or to commit fraud.

See, that's the incentive. Researchers

commit fraud because they are rewarded for publishing often and for publishing big results and for appearing to be geniuses. So they they fudge the

geniuses. So they they fudge the numbers. They sometimes outright lie and

numbers. They sometimes outright lie and they fabricate data. Other times you plagiarize where you just steal research from other people. In other words, you find a way to cheat. I think when students decide to cheat in college,

essentially what they're trying to do is game the system. They want the economic or lifestyle benefits of an education, but they don't want to have to put in the work. Human beings are often lazy.

the work. Human beings are often lazy.

They look for shortcuts. It's a normal way to behave. It might not be an admirable way to behave, but it's at least normal. And the academics who

least normal. And the academics who commit fraud are doing roughly the same thing. They want the benefits of having

thing. They want the benefits of having done this great research, but they look for shortcuts. You look at both of those

for shortcuts. You look at both of those cases, there's one common lesson simply that incentives matter. How you

structure an institution matters and how an institution is structured and what it rewards matters. There aren't going to

rewards matters. There aren't going to be quick fixes to either problem. You

know, part of the problem is you have to convince people that being truthful or doing your homework or really thoroughly going over all your data matters. And

essentially what you're doing is convincing people to spend more of their time doing things that probably they don't really want to do. And and this is I think the big challenge that universities around the country are

going to be facing. I've talked many times in this channel about the ways that universities have sort of lost their way. They are beholden to a

their way. They are beholden to a certain economic model of education which can make it difficult to see why actually getting an education in the first place matters and we're seeing more attacks on higher education from

various parts of the government. A day

or two before recording this, I saw the news that Indiana University, which is a premier state university that's known for its language programs, just cut many of those programs due to low enrollment.

There's a new state law that says that if you don't have a certain number of graduates per year, the major and the PhD program and everything just go away.

And and so now they've just gutted some of the best language departments in the country. When it comes to education in

country. When it comes to education in the United States, I think we're at an inflection point. Universities had to

inflection point. Universities had to figure out what they want to offer to the world, what they want to offer to students and what they want to be going into the future.

People might know that I used to be an academic and one of the most difficult things about leaving academia and becoming an independent researcher was that I lost access to institutional libraries and that means I don't have

access to things like scholarly journals. So sometimes I end up going

journals. So sometimes I end up going down rabbit holes on the internet trying to find the right PDFs. And whenever I do that, I want to make sure that I'm using a VPN. And in particular, I use Private Internet Access, and they're

sponsoring this video. A VPN essentially is a way for you to browse the internet more securely. It hides your IP address

more securely. It hides your IP address and safeguards your internet connection through an encrypted tunnel. This way,

it shields your digital life and keeps your personal information from people who might want to exploit it. Think

people like hackers or internet service providers or even governments. On top of that, many services and websites have content that can only be accessed if you have the right IP address. Basically,

you have to be accessing it from a certain country. There are shows or

certain country. There are shows or movies you can't watch on Amazon or Netflix or the BBC unless you're accessing it from the right place. I

have a hard time watching, you know, the latest Louis Thorough documentary from the BBC because I'm not in the UK. A VPN

lets you choose your digital location.

In fact, with PIA, you can choose from 91 countries. And in the US, they have

91 countries. And in the US, they have servers in all 50 states. PIA works with all major streaming services and they are one of the few VPNs that supports P2P file sharing and you can use it on

Windows, Mac OS, Android, Linux, and iOS. With a single subscription, I am

iOS. With a single subscription, I am able to protect all of my devices at the same time. PIA is the world's most

same time. PIA is the world's most transparent VPN provider. They never

record or store user data, and their no logs policy has been proven multiple times in court. And for an extra level of confirmation, this was confirmed by a third-party audit from Deote. If you use

my link, viewers can get an 83% discount on PIA. That's just $2.3 a month, and

on PIA. That's just $2.3 a month, and you get four bonus months for free. You

can use the link in the description, or you can use the QR code that's on the screen right now. Signing up is risk-free, allowing a 30-day money back guarantee, and they have 24/7 support available to answer any of your

questions.