This is how I'm using LLMs for learning. I've always found on-line classes to be terrible, they're always just a recorded video with maybe a TA to answer questions on a message board. It always feels like it's assumed you're left to your own devices to figure out lessons.My hope is that the initial use of LLM was simply due to people doing one prompt searches (transferring google searches) , as we are already seeing LLM natives using the technology on a more interesting way: upload your manual into it and query it on the friction. I can see this working very well on subjects like maths and chemistry for example, where the student can interrogate the text book on the topics that are challenging to understand and then ask it to generate exercises similar to ones they failed. Or talk them through how they got it wrong.
Yes basically every student can have their own teacher, standardised by the quality of the textbooks.
There are plenty of non-extreme accounts of the rampant cheating; I have high school-age children who fit the mold of today's average teenager and, while I don't dispute your own experiences, they attend a great high school in a great area and it jives.Between a covid generation and LLMs it's definitely been tough, but that subreddit is extreme. It rewards the more extreme stories with more engagement and upvotes so there's a feedback loop where you get the wildest stories.
I used to read that sub but it was just so divorced from the reality we were experiencing at our son's school, one of the lowest ranked, low-income area elementary schools in our county which nevertheless has altogether been a positive experience for our son. I've chaperoned plenty of field trips and events during his time there and the kids are regular kids.
I have to wonder about that, what if you had a friend who took the class and kept his copy of the test? No rule against that (assuming the teacher returns it). Surely just reviewing the prior year's copy of the test isn't in and of itself cheating? If you went to an upperclassman and asked for help and they pulled out their old copy of the test to review that would be fine.One of them had a test bank that basically got the entire group through the first two years of the program at the top of the class. Was it wrong of them to use it? Probably. I consider it “more wrong” that the faculty was repeating the exams verbatim for years. They surely had to notice a disproportionate amount of 100% results!
Hell, I had to do that as a TA. Altering tests so that wrote memorization can't get a passing grade is like...half of the work.You'd have thought that it would set off a bunch of red flags, for sure. Then again I saw exactly the same thing in some of the undergraduate courses I had to take. I'm not against reusing some questions, and I'm certainly not against students having access to copies of past exams. But recognizing that and taking it into account should be a key part of any instructor's job.
Depends on the student, on the teacher, and on what the goals of the student are. ISo what do yall do about this problem? The current situation isn't equitable either because it puts the students who don't want to cheat at a disadvantage.
I mean even those of us who did the work probably did it to launch our careers not because we have that deep a love of the subject. And I don't think it's unique to higher education, I don't think a plumber would go around fixing pipes for the love of plumbing if they weren't paid for it. If you got your Star Trek job and you'd still be there trying to cure cancer in a post-scarcity world good for you but that's probably 0.01% of employed people.Sadly, this isn’t true.
Economically, a degree has conferred better employment opportunities and consequent better pay and quality of life—even better partner selection possibilities. Given this reality, education for its own sake is not the median student’s objective; the goal is an impressive degree as a door wedge to greater economic outcomes. (...)
I was thinking less of cheating and more some of the other extreme social issues you see brought up there.There are plenty of non-extreme accounts of the rampant cheating; I have high school-age children who fit the mold of today's average teenager and, while I don't dispute your own experiences, they attend a great high school in a great area and it jives.
YMMV.
That's a lazy excuse. No company cares about your GPA if you can do the work, get along with others, and add value--or certainly no company I would choose as an employer, a vendor, or a customer.I don't disagree at all about the author's take on LLM use. But I do have some thoughts on the students seeing the instructor as an adversary and the goal being the grade. This is simply a result of how we have chosen to do education.
It's been several years since my school days but I can tell you that I always had two goals in school - to obtain knowledge/skills and to obtain good grades and let me tell you the two have almost nothing to do with each other. Well, that's not completely fair - the author is right about the friction and that the process of doing the schoolwork well enough to obtain the grade would sometimes result in acquiring or cementing knowledge or skills. Emphasis on "sometimes."
The problem is a bad grade has consequences. You fail the class or your record looks bad in a world where you need it to look good. Instead of an opportunity to learn from mistakes and improve, a bad grade is a millstone that follows you throughout your academic career and potentially even into your actual career. Some instructors will give you the opportunity to redo the work consequence-free to obtain a better grade and this is wonderful but it's still burdensome because the class doesn't stop to wait for to catch up and you still have multiple other classes as well as any life responsibilities to keep up with. There simply isn't time to do every assignment multiple times.
So yeah, I cheated. I cheated ferociously. Any chance I got to cheat, I took it. I'm not defending that as the correct moral choice or the responsible one, merely the pragmatic. It's almost certain that this had deliterious consequences on my learning. I'd like to think I wouldn't use an LLM to do my work for me if I were going to school today but I probably would. I should qualify that to say that I probably would if I were taking a full time student course load with the goal of obtaining a piece of paper that states I'm qualified to work in my chosen field. If I were taking classes for no other purpose than to learn and develop skills, I would keep my course load light and I wouldn't feel the need to cheat.
I think we should just look back to how we did this 25 years ago. We all used Sparknotes, and most problems asked by professors were answered in the spark notes. You could literally copy/paste whole sections and just move sentences around and change some words and get a good grade. Fraternities and academic societies have maintained “test banks” for years.
Which is all to say, this is probably not as big a problem as it seems.
Yah, one big difference is that in the before times cheating and not learning the material would eventually catch up to you. Those pre-med students still have to take the MCAT. Students in science and engineering would be learning more things that build off of what they learn in lower level courses and would fall further and further behind. You can find answers to physics 1 and 2 problems all over the place, but get to higher and higher levels and eventually they're screwed. And so you could say, well, tough luck guys, should have studied more in the earlier classes.I've been a physics professor for about 25 years. In that time, I have seen homework scores gradually go up, while students coming to my office for help and average exam scores have gone down. The unfortunate consequence is that exams have become a larger percentage of the course grade. I try to pound it into them that physics is like weight training - watching someone lift weights on Instagram is not going to make you stronger. Those that don't listen fail the exam. As a remedy sometimes I make them explain to me how they did one of the homework problems they got 100% correct, and grade the explanation. That has helped a lot - more students come to my office before the homework is due now just in case. Plus, it makes them realize "Hey, I understand things better after talking with the prof! That actually helps! I might go more."
I'm fortunate enough to teach mostly in-person classes small enough for this to be practical in. And I teach at least some students who actually want to learn the material. I teach a lot of pre-meds who a) know its going to be on the secure MCAT, and b) know it will be of relevance when they actually get to med school and learn about torque in the limbs and conduction in the nervous system. God help us when someone gets an MD by having AI do all the work!. I'm also fortunate that LLM's still aren't that great (yet) at solving carefully worded physics problems. Even when they try to use AI for homework it's pretty likely to hallucinate an answer. There's no question its a problem, though. My exam scores on average are still not what they were 20 years ago. Even without LLM's some of them will just pay Chegg. And I have caught one student paying a person in Asia to do the online homework - they weren't smart enough to submit the answers themselves, so they all came from an Asian IP address. Students not doing the homework is not a new problem. AI has just made it much worse. Getting the students to understand that learning is the point of homework is the only partial solution.
GPA often determines starting salary for new-grads new hires. Not just +/- a few percent, but like up to 30% or more.No company cares about your GPA if you can do the work
This is exactly what I was going to say. What's the point of trying to stamp out the LLM cheaters? You can't force people to learn (that's part of what makes education different from training). Time wasted on a futile crusade against LLM cheaters would be better spent on supporting honest students.
The article ignores the other side of this and conveniently levels the blame on students. That is, instructors are using LLMs to automate their grading. And you can bet your ass the students know it. They are then asking themselves if my teachers can't be bothered to do their job, why should I?
The answer to all this of course is to not to throw up your hands and proclaim its hopeless. But to do what teachers have always done when faced with technological disruption, adapt.
The problem with LLMs is that they take the old "they wouldn't print it if it weren't true" argument and amplify it by a million. Not only are they untrustworthy plagiarism machines, they also tend to lead people to abandon critical thinking and assume that the LLM output is correct. (It often isn't, as you well know.)
I have to say this drove me nuts in undergrad, many of the professors had rules in their syllabus that they will treat having or studying a prior year's copy of a test as academic dishonesty, fail your test, etc. Yet the more warnings against doing it, the more likely the final exam was the same for the last 8 years, and that many of students would have an old copy.Surely just reviewing the prior year's copy of the test isn't in and of itself cheating? If you went to an upperclassman and asked for help and they pulled out their old copy of the test to review that would be fine.
So if reviewing the test is fine, but you knew the questions would be the literal same, what are you going to do? Study it but don't commit any details to memory?
The fact that the questions were unchanged and the professor surely knew what was happening seems like borderline malfeasance on their part.
Amen. I remember drawing all the op-amp circuits on my TI-83 for Intro to EE. Pretty sure I learned them really well just from this level of effort to "cheat". Okay so that probably wasn't your exact point was it....Green could just as well have been describing the process of learning. If there’s no friction, no effort, then no work occurred, and the student hasn’t learned.
This just seems so weird to me, because in the UK my college would sell you any past paper you wanted for essentially the cost of the photocopying. Past exams would give you a really good idea about the kind of question that could show up in the current exam but you'd never be able to just copy/paste answers - especially not from the previous year's exam which was pretty much the worst past paper to go off.One of the most devastating academic disillusionments I experienced came from those test banks.
I was in grad school where ostensibly all of this material is relevant to our ability to perform in our career. I put so much effort into studying for an exam. I’m reading the text, doing practice problems, reviewing lectures, etc etc.
I show up on exam day and there is a little pod of guys I’m friends with reviewing a printout. “What’s that?” “It’s last years exam!”
Every single one of those guys got a 100%. Not a question had been altered. They didn’t study at all, just brute memorization the morning of the test. I got an 85%. And our grades and class rank directly affected our ability to specialize.
One of them had a test bank that basically got the entire group through the first two years of the program at the top of the class. Was it wrong of them to use it? Probably. I consider it “more wrong” that the faculty was repeating the exams verbatim for years. They surely had to notice a disproportionate amount of 100% results!
The one study I've seen on this found that LLMs did pretty well on first-year science content but, by third year, were basically failing.Yah, one big difference is that in the before times cheating and not learning the material would eventually catch up to you. Those pre-med students still have to take the MCAT. Students in science and engineering would be learning more things that build off of what they learn in lower level courses and would fall further and further behind. You can find answers to physics 1 and 2 problems all over the place, but get to higher and higher levels and eventually they're screwed. And so you could say, well, tough luck guys, should have studied more in the earlier classes.
Do you think that's not true now? I bet an LLM could fart out some mostly ok answers for lower-level courses, but if I grab my old copy of Wangsness' EM Fields it will very quickly not be producing believable answers.
Of course I'm no longer smart enough to answer those questions either so it'll be hard for me to test it!
Though I avoid social media, your statement says SO much about why it is a plague"Goofing off with friends" is building social connections in the community. Scrolling Tiktok isn't, but it feels like it is when you're doing it.
Um - TeX is from 1978, the exact same year as WordStar. I believe it had reached the PC by the time your AT&T PC6300 launched in the mid 80s.all this was pre word, wordperfect, teX, and lateX. i was using wordstar at the
time for basic text processing.
Well, obviously this is so. If you write a new test every year, there is no point in issuing warnings against studying off of the old one.I have to say this drove me nuts in undergrad, many of the professors had rules in their syllabus that they will treat having or studying a prior year's copy of a test as academic dishonesty, fail your test, etc. Yet the more warnings against doing it, the more likely the final exam was the same for the last 8 years, and that many of students would have an old copy.
I believe LLMs are great at these things and become better everyday. (I know that's a minority position here.)My hope is that the initial use of LLM was simply due to people doing one prompt searches (transferring google searches) , as we are already seeing LLM natives using the technology on a more interesting way: upload your manual into it and query it on the friction. I can see this working very well on subjects like maths and chemistry for example, where the student can interrogate the text book on the topics that are challenging to understand and then ask it to generate exercises similar to ones they failed. Or talk them through how they got it wrong.
Yes basically every student can have their own teacher, standardised by the quality of the textbooks.
the medical profession understands how useless credentials alone are. that's why they have the worlds heaviest apprenticeship program.You can't just test a surgeon and let them loose on a patient unsupervised
It's a real problem, but I feel like it's touching on another one at the same time. There's a paragraph towards the end where the author notes that no student using a LLM actually thinks they're learning, and that they're using it as workload management.
This was true in my time as well, but it's only become worse since: students are doing triage. The old cliche goes, "Will this be on the test?" And it's an old cliche.
Yes, LLMs in education are a blight, but their use by such a large proportion of students is a symptom of the larger problem. Students, like everyone else, have to decide how to spend their time and effort to get the most out of them. Time management and prioritisation like that is a skill, and one they tend to be taught late or not at all. This while contending with authorities who usually don't have their best interests at heart, and will quite happily lie about what's important for the student to suit their own agenda, or more likely just not consider the student's needs at all.
This is not meant as a criticism of the teachers on the ground, like the author. They're the ones most likely to actually care, if that system hasn't worn them down too.
The one lesson every single student learns early is that the school as an institution is not there to help them. It's an obstacle course they have to navigate. They all want to have a good life for themselves, every single one of them, but they have to try and judge what will actually help them to achieve that, and what's only there to make them a more valuable commodity for someone else at their own expense. And since most of them are young, either very young adults or literal children, they don't have the experience to avoid making big mistakes when navigating that minefield.
They're using LLMs as a tool to try and manage the parts they've identified as unimportant or detrimental to them. Like the author, I think that, in itself, is one of those big mistakes. But I also think it's critically important to be aware of why they're making it.
Well said and well thought out, I agree with pretty much everything.I work in academic technology at a college and I'm currently both doing focus groups with students on AI use as well as designing a series of summer workshops for faculty on AI in the classroom.
The most salient comment I've heard in the past year was from a frustrated student who works hard on papers only to see other students get a higher grade when they spend 10 minutes with ChatGPT because "ChatGPT gives the answers faculty want"
This reveals a serious disconnect between what faculty want (student learning) and how they assess that. We default to grading term papers because those were reasonable instruments to see student capabilities- we don't actually care about the term papers themselves. (We're more than happy to throw then in the trash after a long weekend if grading). We just want to know if a student actually understands what we've been teaching
This is not a new problem- students have always been able to bypass writing papers by getting other (paid) people to write them -LLMs just make that easier and cheaper. The answer isn't to give up or to add some ridiculous amount of anti cheating tech, it's to change how we assess learning. Does that mean in-class handwritten assignments in a blue book? Maybe, if that's appropriate. Oral exams? Sure. But get more creative- I've used serious role playing games where students reenact historical events like the trials of Galileo and interactive simulations to let them play around with how an industry changes over time. One of my guys runs our innovation lab (makerspace) and he's amazing at figuring out ways to get hands on activities mapped to learning goals.
Is reworking your entire syllabus to deal with AI going to be a pain? Yes. Is it going to take more time? Probably. But it's not impossible to figure out new ways to assess learning- and who knows, they might be a lot more interesting than grading yet another stack of mediocre term papers