Like other comments, I can't help feel uncomfortable with the way everything was written defensively, where every example is used to justify the author's actions.
The author concludes near the end:
> That said, I have no judgment to pass on you here.
And follows with this as the very next sentence:
> We are all human beings, we all make poor choices and mistakes in our lives.
Which seems to be clearly judging the students for their "poor choices and mistakes".
I feel the author is out of touch with the students and the times; If so many of them are leveraging AI perhaps the better move is to find ways to align that usage with learning. He quotes:
> "it is better 100 guilty Persons should escape than that one innocent Person should suffer"
But if there is such a scenario in a classroom rather than a courtroom, then perhaps the methodologies should be revisited?
It feels like the author is the ACKCHYUALLY meme, striving to prove themselves technically right, rather than empathizing with the students, which I feel shows a lack of maturity for an educator more than 2 decades older than his students.
I understand that education and evaluation is difficult in this AI era, but I can't come to agree with the author's actions or conclusions in this case.
With regard to judgment, there is a difference between judging a person's character and judging their actions. That was the point there. The actions clearly violated course policy.
I am curious what leads you to believe that there's a desire for someone to agree with the actions when the retrospective opens by saying the situation should have been addressed better.
Assuming that the students were using generative AI when told not to...
The way you leverage AI for learning is to do the work yourself and ask it to review and provide feedback, or to ask it to explain some concept, ask it questions to check your understanding, ask it to generate questions for you to answer, etc. If the entire point of the exercise is the exercise, then generating the answer... misses the point.
Like I can easily find online how to encode if statements in the lambda calculus, or how to make a half-adder out of logic gates, or how to prove that bijections are invertible. It's always been cheating to not do it yourself. School intentionally gives you easy problems with known solutions, not open research problems where you should use any and all tools and information you can find.
> If so many of them are leveraging AI perhaps the better move is to find ways to align that usage with learning
Should the course be updated to better leverage AI? Sure.
But as a student, if you're explicitly instructed not to use AI for a particular course/assignment and you do it anyway, that's academic dishonesty. No matter how much you disagree with the principle of the thing, you still lied and cheated.
I'm pretty convinced now that grading take-home assignments is pointless with the prevalence of AI cheating. (There's surely already "innovation" going on targeted at letting people cheat on assignments while staying undetectable.)
It's probably best going forward to grade only in-person proctored exams using paper, or offline air gapped computers in the case of programming courses.
Or don't do exams at all, focus on projects that are hard even with AI. Grade on readable and well structured code, performance and UX instead of just following the assignment point by point. The problem is just that it adds more work to teachers.
The most practical solution I know is the teacher asks you questions about your code in person to make sure your understand it. Even better when they come up with a small change and you have to implement it on the spot, or at least explain exactly how you would do it if there's not enough time.
It catches lazy vibe coders who don't even read the code but lets through those who only use AI to save time. That's exactly what you need if you want your students to learn.
We already know the solution. What did colleges have students doing to ensure they knew how to use microsoft office? Three 50min lectures and two in-person 1.5hr labs a week for 2 credits, all dedicated to using word and excel. For a normal person in the 2000s, this is ridiculous to force on all students. It was not ridiculous 10 years before. Today, students again need the forced in-person learning for office applications.
For CS, homework should be done how you want. Homework is practice. Let AI assist you, but you better learn it because you'll be coding some part of it in lab with a physical programming book as a reference. They can lock down computers for labs to prevent web access or limit only to certain websites. Since labs are weekly, it will be obvious if someone is not learning as they do the homework. It won't be an F on the final as the first sign of trouble.
Schools can do things to improve student resources. Self host llms that are not corporate controlled that every student can use. The school can provide a search engine that is useful and not google's marketing. You can offer an education proxy that caches the pages you look at when doing education work, preserving anything you cite. This can all feed into the school llms for rebuilding to keep them trained on current info.
I had many classes in college that didn't administer homework or take-home assignments. All we had were just exams. Most of those professors operated on the logic that if you wanted to pass their class, then you would do the (suggested) work.
Also, I never had a computer science class where exams were administered on a computer. I only graduated like 10 years ago, too. I shit you not, even my x86 assembly class was all pen and paper too, and yes, we were graded for code accuracy.
* Do not use variable length array in this class
* Some of are you are copying code and concepts from things outside of course material
* `fgetc()`, `sizeof()`, `malloc()`, etc
* You're probably cheating.
* You're also making your life more difficult
I always found these sorts of instructions quite frustrating as a student. If I'm a student enthusiastic about learning a language, I would have definitely been one of the kids trying to figure out how to use malloc and creating novel approaches to the projects that are assigned to me. It's a university class, why the hell not? Besides, isn't that going to be an advantage towards me considering the code is going to be evaluated by a anti cheat engine?
The AI stuff aside, this comes across as a professor who feels threatened when his course isn't taken as doctrine. Being called a cheater for using external resources is ridiculous to me.
Yeah, and sizeof and malloc are very easily things students could have been exposed to in high school or their own studies. I’ve seen C books that use them in the first chapter.
It’s fine to say solve this problem using the following tools, but not great to say “if you find another approach, I assume cheating.”
It did not say "I assume cheating". It said you're probably cheating, as in of you look at the population of students who are using such constructs, over 50% of them are using them because they are cheating. Given that accusations of cheating were not made when such constructs were found, it seems like that probability was not factored in when deciding whether or not to advise someone.
It does not say that one cannot use those things, it says that one is likely making their life more difficult. Implementations that utilize those things are, with high probability, more convoluted and involved. This slide is specifically around the time we introduced fscanf() and friends. If one is using the things mentioned in the slide, they're likely partially re-implementing *scanf() on their own. Which, while certainly a good exercise, misses the goal of developing familiarity and understanding surrounding *scanf().
While there are certainly a handful of experienced students that use these things, ten years+ of experience shows that most are copying from elsewhere. A simple, brief meeting with the student is easy enough to ascertain that, however. And that was certainly a part of the process.
Regardless, that alone would not have triggered an academic integrity investigation and, again, there was no prohibition here.
As a student, _especially_ in the weedout classes where your performance in those classes decided if you could get into that major (I went to a school where you applied to majors 1-2 years after starting) it was incredibly frustrating to see other students cheating. When applying to a major that was well past capacity and very competitive to get into, it very much felt like a 0-sum scenario where cheating students were screwing the rest of us over. How awful would it have been to be declined from your major and have your career path severely impacted because so many students were cheating with 0 consequences
It was even worse when professors were notified but couldn't do anything about it (for example, a student next to me in an exam kept pulling out his phone but hiding it when the professor was nearby. I even notified her during the exam but she wasn't able to catch him so as far as I know, it never went anywhere.)
Similarly, I was a grader for a 300 level math class where about 50% of the class had blatantly cheated on an assignement (the solution manual on chegg had mixed up the questions so anyone who actually read the problem would've caught it because it was the wrong problem). I notified the professor and he said he would mention it in class but no other consequences occurred - they didn't even get 0s on the assignment!
All of your thought processes were very reasonable and I thought the best slide was comparing exam scores by people using LLMs on assignments vs people not using LLMs on assignments. I truly think that will convince people who otherwise might be borderline failing and desperate to pass that using an LLM is not in their own best interest.
I'd be curious to see (both as myself, and I think it'd be very convincing as a student in your class) the overall class grades for people using LLMs vs not. I could see someone reasoning "ok well if I get 100% on the assignment and scrape by on the exams, that still gives me a better grade than doing badly on the assignments (without an LLM) and doing slightly better on the exam"
Edit: I always thought for some reason that cheater would get their karma in their own way but in the real world, there are no consequences. I'm plagued by ads for ai-assisted ways to cheat on interviews every time I login to Blind. It actively is making things worse for the rest of us
Maybe you would have. Maybe you wouldn't have. But they have to design these things for the 90% case. People using things that are far outside what's being taught—in the age of AI—are maybe more likely to be cheating than everyone going above and beyond.
Students are human. Many have been taught to focus on one end goal: a grade. They see a mythical program that knows 6000x more than is taught in this 200 level course; of course they want to use it!
This isn't the 90% case. At least half of CS students come in with at least a basic proficiency in programming. It's a big enough portion of the class that not boring them into ignoring the actual Computer Science parts while getting the other half up to speed is the defining issue of lower-division CS courses.
When I took my Freshman year undergrad programming course decades ago, I had already been programming in that language for at least three years. It would have been quite painful to have to deliberately not do things I already know how to do because "I didn't learn it yet."
This was definitely a problem in classes before AI. I had a number of infuriating teachers who insisted on failing me for doing X, because "you're going too fast and we haven't gotten to X technique yet."
Decent professors distinguish between "using stuff we haven't taught yet" and "trivializing the assignment by using stuff that wasn't intended". If the assignment in an Algorithms course is to implement a square root function, calling "sqrt(x)" is obviously inappropriate. Structuring their solution using a do while() loop rather than while() loop is not.
I remember being marked down for using a ternary in Java for an introductory programming course at uni because the teaching assistant didn’t know what it was. Pretty minor loss of marks but it always annoyed me.
That said, having to do the introduction to programming class at uni despite programming for like a decade in high school was the bigger annoyance.
Same. I taught myself C when I was in high school, back in the early 90's. Fortunately, in college, I was able to test out of the intro classes for credit.
Are you so graybeard that structured programming was interesting enough to be part of the focus of a C course? Because I can see in that context requiring the use of appropriate control flow structures. Nowadays, I can't imagine in what context students would be so eager to reach for GOTO that they don't preferentially use other control structures.
I'm surprised by the negative response towards this professor. The syllabus was extremely clear up front, and students did not respect it and decided to see what they could get away with instead. Yeah there's some weird rules to follow, but the homework seems fairly scoped to smaller problems and it's just one class out of a few dozen.
Because it doesn't matter if he is right or wrong. A bunch of kids failed an important class and he didn't do a damn thing to address it. The teacher has failed. If you were his employer, would you think his methods are effective with the current students? I don't. Kids are lacking so many experiences, its just not reasonable to expect self study on everything. It also isn't required for learning.
I don't get why this is so hard to address. Make the homework a small part of the final grade and use in-person labs to ensure students are learning how to do the homework. Take advantage of llms as a tutor and helper, it is silly not to use the tools we use for so many classes already because we want kids actually doing the work. No one needs to shine light through a prism or swing a pendulum to pass phyiscs, they have labs and still make you do it.
In a lab, you can prevent the use of an llm and control the websites accessed. You can even put a relevant programming book at each station. You can do individual work where people all work alone and don't talk or you can do group work and vary during the semester. Make it a 3hr lab twice a week if that is what it takes. Phones stay in bags or pockets with dnd set.
I also think schools need to provide in-house llms for students to use. Labs, homework, research, etc should all be done using llms hosted by the school. It's crazy if this isn't already happening at every major university.
Why give all that student data to a corporation using it to make your students jobless? If you keep all the search and llm use in-house, you can use the data for education improvements and research open to students.
Anyone who is lame enough to cheat on an intro to C class is doomed anyway; the AI they cheat with will be the AI that replaces them as a worker because they didn't learn anything. Give 'em hell, Prof. Turkstra!
Professors will be the last to admit their skill set is outdated. Those folks and academia in general aren't outcome-driven. They don't care if students will be ever able to find a job or not. They just want to educate like they got used to, and want their money. Or is it me being jealous? I had to change. Change a lot, learn a lot in last couple of years. But what about those folks who still use Borland compiler? I can imagine they're comfortable in academia, and can be in the next 10 or even 20 years - depends on how aggressive they're defending their old-world principles.
I was an assistant teacher at a private engineering school, teaching a toy kernel project.
We had to add a new significant assignment which was the student sitting down next to me and we both look at their code and I ask questions.
Our policy for the project wasn't even anti AI, it was simply "do not submit code you do not understand" so obviously we had to test understanding, but already with 90 students it was hard, I don't know how you would manage something like this with 500
> This form must be completed .... Failure to complete the form will result in an unfavorable letter being sent to the Dean of Students along with further potential disciplinary action.
> Through careful analysis and manual review, we have identified what we consider to be clear and concrete indicators in one or more of your homework assignment solutions that it was partially or entirely generated by an AI/LLM tool like ChatGPT, Claude, Copilot, etc....
> You are required to submit a response to this form....Failure to respond will result in a grade of 'F' for the course.
> Honest responses that align with our analysis will result in a score of O for the identified assignment(s).
> You also have the option to meet and dispute our findings. Should you choose to do so and we determine that our original findings of AI/LLM usage stand, we will instead apply the standard course policy from the syllabus and assign a grade of F for the course.
The above letter sent to a bunch of students.
> It is an open question whether this is reflective of their actual behavior or instead something that underscores the purported, and unintentional, potentially coercive nature of the form.
Sounds like the professor discovered (or disputes the possibility of?) the concept of forced confessions.
"Honest responses that align with our analysis": did the professor consider the possibility of dishonest responses that align with the analysis?
"Should you choose to do so and we determine that our original findings of AI/LLM usage stand" Not much needs to be said about that.
If the professor was confident enough about the analysis to send the letter, then all of the students getting the letter should have gotten an F and no letter at all.
> If we find reason to believe that a student or team has cheated on any assignment, we may inform the student or team promptly, or we may decide to silently accumulate evidence against the student or team on later assignments.
Why role play what you may or may not do if you encounter cheating.
The university has rules about how academic integrity is handled independent of the syllabus.
Including this line in the syllabus, and again at the top of the blog post is weird
That's not roleplaying, that's the basics of how you handle stuff like this. For lower-division level technical assignments, there's only so much free expression to extract signal from with regards to copying per assignment. Any individual assignment could likely be appealed as coincidence. If the same students turn in 2 more assignments that are suspiciously similar to the same other students, then you can have a pretty iron-clad argument for cheating.
Students cheating at lower-division university level courses is almost entirely a symptom of the students being poorly suited for the course material. Students who are interested and capable, but overwhelmed with the amount of work tend to just turn in incomplete assignments, or not turn in work at all. Encouraging the cheaters to drop while they can still get "dropped" rather than an "incomplete" or "F" is in everyone's best interest, as it provides an opportunity to consider a pivot to something that interests them more.
IMO, the overreaction by the professor was seeking further penalty for students who already dropped the course, or who would do so when confronted with their misbehavior. A student who cheats in CS is not necessarily a student who will cheat at chemistry, or business, or journalism. If they've already dropped out of CS, there's little benefit to be gained by pursuing the matter further.
> that's the basics of how you handle stuff like this
Anecdotal but not in my experience. The professor can inform students, collect evidence, etc without detailing each scenario in the syllabus.
Every CS class I’ve been a TA for had the same generic university provided statement on cheating.
The professors did not want to spend time thinking about or dealing with people cheating.
Every time MOSS flagged assignments they basically sat with me and the student and said “You are likely cheating. Please don’t do it again.”. No one ever admitted to cheating and no one ever got flagged twice.
I recently completed my masters in CS. I agree with the sentiment of wanting to avoid LLM use in general, but that ship has sailed. No amount of policies and slides is going to be enough for this systemic issue. I remember all the policies and volumes of reading I would have to do to keep up with what was expected to be done in a very rigorous and usually pedantic way. Change to oral exams or something similar
It's beyond me how binary their thinking is when dealing with an issue as nuanced as AI.
For instance euros have oral exams often and they are much more helpful on seeing if the kid has any idea on what they are doing. Furthermore it forces them to think through the problem more clearly since they have to explain it to someone else at some point (rather than some autograder).
It's part of the problem with academia being behind industry.
I had a similar problem when doing my MBA; Lean Startup had just been released and the whole "startup scene" was getting going. But I couldn't use any of that in my MBA entrepreneurial course, because no-one had written a paper on Lean. I had to write business plans (which, obviously, no-one was doing out in the real world). I was being taught a useless skill because academia couldn't adapt quick enough.
For these CS grads who are being punished so harshly for using AI, the same: the first thing they will meet when they start their new developer job is a Claude prompt. Their CS course will not have prepared them for this, and they will suck at using an LLM to generate code.
There is an argument that universities should not be vocational, and should teach people how to be better humans, not better employees. I agree completely with this, but that also means universities need to stop advertising how getting a degree leads to more lucrative employment. Which is it?
Because if it is vocational, and if a CS degree is supposed to lead on to a career in software development, then they need to teach orchestration not C. We don't teach prospective developers how to write machine code on punched cards any more, for exactly the same reasons.
> For these CS grads who are being punished so harshly for using AI, the same: the first thing they will meet when they start their new developer job is a Claude prompt. Their CS course will not have prepared them for this, and they will suck at using an LLM to generate code.
You're assuming that this policy continues for higher level courses. It does not. Use of LLMs is discussed, and in some cases even encouraged, in higher level courses like Software Engineering I.
I would argue that if they use an LLM at the stage of a lower-level fundamentals course, they'll suck even more at using an LLM for more interesting tasks. They also won't have the skills necessary to address problems when the LLM invariably doesn't do exactly what is desired.
That's good to know. I'd assumed that this was one of those courses.
> I would argue that if they use an LLM at the stage of a lower-level fundamentals course, they'll suck even more at using an LLM for more interesting tasks. They also won't have the skills necessary to address problems when the LLM invariably doesn't do exactly what is desired.
That's how it works now, kinda, with all us seniors controlling the LLMs. But we're learning a whole new set of skills that are only tangentially related to the skills required to actually write code. I can easily see that in a couple of years students are taught those new skills directly.
> There is an argument that universities should not be vocational, and should teach people how to be better humans, not better employees. I agree completely with this, but that also means universities need to stop advertising how getting a degree leads to more lucrative employment. Which is it?
I read a reply to a similar question in a different HN thread a while back that stuck with me. It went something like, "The point of university is to get an education. The point of student loans is to get a job."
One of my hopes is that AI screws up the whole career thing so badly that we go back to education being non-vocational (because there's no point, everything they teach will be obsolete before the students graduate).
> Every student is also responsible for protecting his/her own work. If you inadvertently allow your work to be accessed, you may still be held responsible for facilitating an academic integrity violation.
This is insane. Absolute power tripping. And I say that being in favor of banning AI throughout the course.
I'm a fan of zero tolerance policies for academic dishonesty. The dean getting involved and mitigating consequences for the ~~customers~~ students is classic.
> Do not look at other students’ homework files and do not work with other students. At all.
This is a horrific policy. We work better with other people.
When I studied computer science, other students came to me for help. I helped them. We discussed problems together. I saw their code. They saw mine. No one copied: we shared approaches and style and thinking. I shared how I solved things. They learned my thinking. By explaining, I understood what I knew better. Sometimes I learned from them: an idea I wouldn’t have had alone, a new direction. All of us were better off. This was across multiple people.
It was in no way discouraged and so long as you wrote your work — while asking others for help was fine — you were okay.
Humans learn through teaching. We learn through discussion. We learn through interaction. Solitude — ‘do not work with other students. At all’ — is the exact opposite of a healthy learning environment.
I don’t care what their artificial intelligence policy is. I cannot state strongly enough how much I revolt against their human intelligence policy.
As for the lecturer’s warnings in slides: that reads like a trap.
The course Programming in C, and other courses like Compilers, have this policy across many CS universities for many years now. It s a strict policy but it has its reasons. Assignments on early semesters are simple and short, and if students cooperate on the implementation, the result would be almost identical for everyone, while very few would have truly thought it out by themselves. Most students would not go easily through the process to struggle in order to think about the solution, if they could just copy the answer from someone else.
Thanks for the reply. I get that, and maybe I am approaching it from the perspective of someone who wanted to learn (my friends too), not someone who just wanted to pass.
But the situation I described included the basics. I specifically remember explaining pointers to a friend and working through a simple assignment on C pointers with them, for example. Absolute basics, yet collaborative working helped them greatly.
Perhaps your policy puts TAs there in that situation, and I understand that.
It's important to note that the students are encouraged to interact with the many teaching assistants that we have available. And, while that slide is somewhat harsh in how it states it, the larger lecture discussion surrounding it makes it clear that students are permitted to discuss assignments in a general way. When you get into implementation details, that's where the academic integrity problem arises.
For people that haven't spent time teaching lower-level programming courses, policies like this may seem overly harsh. But, remember, they're not solving particularly advanced problems at this stage. Instead they're mastering fundamentals. An analogy I like to use is learning to play the piano. If someone else is doing it for you, you're not going to develop the skill.
The main reason for the policy is that many students are not able to walk the line between leading someone to a solution and giving them a solution. The teaching staff know how to do this. When students help each other in situations like this, many times they end up just providing code or allowing someone to look at their code (ultimately copying it), etc.
I appreciate your reply, and thankyou. It did strike me as harsh, yes, and I was rather horrified when I wrote the parent comment. But I think I was also responding to the tone of the article, which struck me as almost setting traps - that kind of emphasis on the student being responsible for understanding something broad yet not saying what it is specifically in the broad material they need to look out for, for example.
Yet I also understand your view -- students providing and copying code yet not learning and doing work themselves. I completely get that needs to be found. Yet, reading the article I felt truly scared for the students who might get caught up in this yet I don't think would have actually done anything wrong.
> Yet I also understand your view -- students providing and copying code yet not learning and doing work themselves. I completely get that needs to be found. Yet, reading the article I felt truly scared for the students who might get caught up in this yet I don't think would have actually done anything wrong.
I get that, and I think that is partly what drove a lot of the initial turmoil when this started. But, the important thing to recognize (and this could have been better emphasized for sure) is that it's really easy to tell whether someone caught up in this is cheating. A 5 minute discussion is all it takes. And, yeah, maybe some could BS their way through a discussion like that if they had a heads up, but at that point 1) they've now learned at least some of the material and 2) you're not going to catch everyone anyway. Processes like this are somewhat like a police officer catching people speeding. The enforcement alone discourages the behavior, to a degree, and you're never going to catch everyone.
I agree. Here at U. Lisbon, we are following the recommendation by the recent MIT report: Evaluate students by talking with them, making it a social interaction, which they are lacking.
We need students to work together and share their experiences. Then we need each one to achieve the learning goals (and we should not assess work they did far from our eyes. We should assess their understanding and skills in front of us).
I had a CS professor who worked like this. We had assignments and tests, and they did have grades, but your overall grade in the course ultimately boiled down to convincing him that you understood the course material. As someone who excelled at the material, but could be a bit hit-or-miss on bothering to complete assignments, I loved it. Students who were used to optimizing their efforts and obsessing over each point hated it. It didn't come up in that course, but the issue with such a grading scheme is that sometimes students and professors really don't see eye-to-eye. Most of the time it's the professor's fault, holding some sort of prejudice. A grading scheme like I described would be disastrous for such students.
This is a good approach, and is used in some situations. But, it has challenges in terms of scalability. It is not impossible for larger enrollment courses (this course had 600 students), but there are certainly trade-offs in terms of consistency in grading, overall student experience, etc.
Sounds like to me maybe universities should stop focusing on being cash grabbing machines and instead properly teach with a useful instructor-to-student ratio?
Sorry, if your answer as to why using a beneficial pedagogical approach is your class sizes are too large, then it sounds like your class sizes are too large!
learning to think is orthogonal to american universities, who's goal is to get government money via fed backed loans to fund professors real hobby: doin cool science
standardized and written tests, online school, homework, etc is just a way for professors (and increasingly, just TA's) who have no interest in teaching, to get the job done with less time and more efficiently.
Some of us are perfectly happy to learn in silence and solitude, and we thoroughly excel at doing so. Sorry that you don't. The broad sweeping "we humans" language gets a groan out of me. It's the sort of mentality that leads to unproductive gatherings in the working world where people want to force "learning through collisions/synergy".
>> Do not look at other students’ homework files and do not work with other students. At all.
> This is a horrific policy. We work better with other people.
I think that line was in connect with quizzes, which are presumably online. Near the top of the article it says
> You may discuss assignments in a general way with other students, but you may not consult anyone else's work.
I'd be with you if they made everyone work in "clean rooms" but that is appropriate for quizzes and tests.
As someone who's graded lots of undergrads I can tell you that cheating is prolific. It's easy to identify who is copying code, but it is hard to prove. My wife teaches in another department and says the same thing. AI use has grown exponentially and it's really destructive to the education process. Though I have more complex thoughts on this and that it is just exacerbating larger problems that no one wants to actually solve because they'd be very hard to solve in a way you can have clear metrics for the bureaucrats (including industry). Proper evaluation is unfortunately always going to be fuzzy. Metrics are in contention with making a meritocracy, especially if you don't have a deep understanding of said metrics
naive to think people would follow this policy in the spirit in which it’s intended, but even then it clearly selects for the most social animals amongst your students instead of the most talented. And discussion alone is all it takes to describe the right approach, so if that’s the test, it’s meaningless to protect the mechanical syntax of implementation.
Clearly in this world the frats and the sorority members get the highest scores, and after that the best looking, the most amusing or charming etc.. so unless that’s the intent the whole approach is perhaps worth reconsidering.
If you want to teach to improve your understanding the normal thing is you become a TA so that access is open.. not open every assignment to ad hoc collaboration so you disadvantage half of the students on every assignment
It may be surprising to you but, caring about fairness and democratizing access to helpful resources doesn’t actually require any kind of disadvantage or a complex. What’s your deal though, since you want to attack that?
I'll pile onto the agreement stack. Abso-100%-freaking-lutely. I worked with my friends through all our assignments, bouncing ideas off each other, talking through confusing bits, etc., and then doing the work ourselves. We didn't ask "how do you write this line of code", but "how are you approaching this problem?". I learned as much from the discussion as I did from the implementation. Throwing that out is absolute madness. No way.
I sat on a university student disciplinary panel. I assure you that your experiences are biased. There are plenty of people that flat-out copy each other.
GP was describing the platonic ideal of why we have universities instead of just textbooks, so I'd venture a guess that you're the one with a biased perspective here. Sure, some students cheat, and the appropriate penalty is throwing them out of the program. But pursuing that end to the point that you forbid students from engaging in normal collaborative learning is nuts.
As you point out, there are plenty of students who flat-out copy each other. Filtering out those students does plenty to enrich the remaining students, you don't need to go on witch hunts to find ever-more subtle cheating.
Given the professor's professional interest in cheating-catching tools, I'm suspecting wanting to get good, interesting data motivated a lot of their behavior here, even in ways he might not consciously realize.
Any policy that asks students not to use AI and trusts that they will comply is giving a massive advantage to dishonest students and punishing honest ones.
This article sounds like the professor was reprimanded by higher powers and tried to dig themselves out of a hole for having a badly structured grading system.
Students should be encouraged to use AI in their learning, because they will anyway. The understanding should be that they will be assessed in supervised environments so they need to learn the work either way. This gives the students incentive to learn without fear of using the best tools available to do so, and provides evidence to the powers that they actually did so.
If you remove the mention of AI from this post, the authors process and setup of the notice to students is absolutely mental. The intentions might be good, but damn if this isn’t a comically inept way to set up a kafkaesque nightmare rather than a place for actual learning. The defensive attitude doesn’t help either. Just the sheer energy that went into all the steps mentioned here, not invested in any even superficial attempt to perhaps adapt. That isn’t to “give into AI” or cheating, but if 50% of students break your rules, more rules and litigation about the rules are not the next step to take.
My comment explicitly did not focus in the use of LLMs and in fact much of the post doesn’t. Leading with the depressing “work alone” slide, it focused on similarity, not AI use. Between students, ruling even “inadvertent” exposure as the same punishable (and literally literally undeniable!) offence. There’s millions of ways to structure a class for learning, this myopic focus is not one of them. Ironically the setup ignores the very basics of human collaboration. AI merely amplifies extremes, it’s a factor but neither the cause nor culprit here.
The title includes “retrospective” but the learning and changes following such a process are absent.
working in a team is valuable. but that value is proportional to the teammates chosen. teammates who have worked alone and can think through their decisions are more valuable, thus it is this way.
>AI merely amplifies extremes, it’s a factor but neither the cause nor culprit here.
you may have missed this part: ~As discussed in the paper, we have also run the tool against earlier semesters with very compelling and interesting results. In particular, the indicators that we use are virtually nonexistent prior to 2024 and their presence increases rapidly over the following years.~
So it stands to reason that LLM's are in fact, the culprit here.
All these analogies are kind of tortured as they ignore the elephant in the room. The students want to be there to get their job getting paper. They aren’t there to learn.
If society offered some reward for lifting the weight while not enforcing how, forklifts would absolutely be used to lift weights.
The job-getting paper ceases to be meaningful if it no longer verifies competence. If it merely verifies the ability to prompt an LLM then nobody will pay a premium for people have it, and then nobody will pay to obtain one.
Yea, this letter could have been written about garden variety "suspected cheating" 20 years ago. And the professor's attitude is the same as a couple of professors I remember back then:
Prof: "I know you cheated."
Student: "How do you know it, can you prove it?"
Prof: "I have ways, trust me I know."
Student: "I know the material, I'll re-take the test right in front of you if you want."
Prof: "Either admit that you cheated and get a F, or don't admit and I'll send you to the school's academic integrity department."
This is the professor who had beef with a /middle schooler/ for copying some content on how to use a calculator from his website. His defensive attitude is clearly indicative of the massive, massive chip on his shoulder. Just reading this triggered me to my experience with similar professors in college about a decade ago. What a jackass
The insistence that you must "work alone" and that similarity between submissions will get you flagged would make me really uncomfortable. I think I would be afraid to talk to the TAs lest I get poisoned by their knowledge.
I think there are two somewhat orthogonal things addressed here. The first is adapting in the face of AI, and that is happening albeit not instantaneously. The other part is simply ensuring students are working within established policies. In some sense at this level using AI is similar to handing a student a calculator while teaching them to do basic mathematics. They're not going to learn addition and subtraction if they don't actually do the work.
As far as the "comically inept" setup, I think that's acknowledged in the retrospective - including an apology.
Can't they just do in-person testing? Build a cluster of cheap, crappy dumb terminal from 40 years ago. Have students type out the source code live. Grade the work. Hell, let them use reference books! Just don't allow re-typing AI-generated code on notebook paper. They can cheat all they want in the interim, it'll all become clear on the midterm and final exam.
This sounds snarky, but I don't quite get why. That's exactly what was done, and in the end nobody needed to be (nor was) removed from the building. Although those not enrolled in the course were relegated to the surrounding open spaces/hallway. Which was a good thing, because lecture attendance on that day was near 100% (i.e., almost all seats were occupied by students).
it's hard to rein in cheating with LLMs in the classroom. this guy is trying at great cost to his personal life - i mean, if not in academia, where else are mean people supposed to achieve great things?
imo watermarking will do little to reduce this awkward, personal, adversarial thing. as always, technology cannot 100% solve, or even 50% solve, a political problem.
> the decision that was reached was to reverse the retroactive use of the tool on earlier homework submissions... Students that had dropped the course would be permitted to re-enroll.
it's too bad the professor completely capitulated. like people have so much angst against H1Bs, so much excitement about unionizing, agitation about layoffs or whatever... look, 40% of your classmates are cheating. 40% of your colleagues are cheats. the call is coming from inside the house!
Like other comments, I can't help feel uncomfortable with the way everything was written defensively, where every example is used to justify the author's actions.
The author concludes near the end:
> That said, I have no judgment to pass on you here.
And follows with this as the very next sentence:
> We are all human beings, we all make poor choices and mistakes in our lives.
Which seems to be clearly judging the students for their "poor choices and mistakes".
I feel the author is out of touch with the students and the times; If so many of them are leveraging AI perhaps the better move is to find ways to align that usage with learning. He quotes:
> "it is better 100 guilty Persons should escape than that one innocent Person should suffer"
But if there is such a scenario in a classroom rather than a courtroom, then perhaps the methodologies should be revisited?
It feels like the author is the ACKCHYUALLY meme, striving to prove themselves technically right, rather than empathizing with the students, which I feel shows a lack of maturity for an educator more than 2 decades older than his students.
I understand that education and evaluation is difficult in this AI era, but I can't come to agree with the author's actions or conclusions in this case.
With regard to judgment, there is a difference between judging a person's character and judging their actions. That was the point there. The actions clearly violated course policy.
I am curious what leads you to believe that there's a desire for someone to agree with the actions when the retrospective opens by saying the situation should have been addressed better.
Assuming that the students were using generative AI when told not to...
The way you leverage AI for learning is to do the work yourself and ask it to review and provide feedback, or to ask it to explain some concept, ask it questions to check your understanding, ask it to generate questions for you to answer, etc. If the entire point of the exercise is the exercise, then generating the answer... misses the point.
Like I can easily find online how to encode if statements in the lambda calculus, or how to make a half-adder out of logic gates, or how to prove that bijections are invertible. It's always been cheating to not do it yourself. School intentionally gives you easy problems with known solutions, not open research problems where you should use any and all tools and information you can find.
> If so many of them are leveraging AI perhaps the better move is to find ways to align that usage with learning
Should the course be updated to better leverage AI? Sure.
But as a student, if you're explicitly instructed not to use AI for a particular course/assignment and you do it anyway, that's academic dishonesty. No matter how much you disagree with the principle of the thing, you still lied and cheated.
I'm pretty convinced now that grading take-home assignments is pointless with the prevalence of AI cheating. (There's surely already "innovation" going on targeted at letting people cheat on assignments while staying undetectable.)
It's probably best going forward to grade only in-person proctored exams using paper, or offline air gapped computers in the case of programming courses.
Or don't do exams at all, focus on projects that are hard even with AI. Grade on readable and well structured code, performance and UX instead of just following the assignment point by point. The problem is just that it adds more work to teachers.
The most practical solution I know is the teacher asks you questions about your code in person to make sure your understand it. Even better when they come up with a small change and you have to implement it on the spot, or at least explain exactly how you would do it if there's not enough time.
It catches lazy vibe coders who don't even read the code but lets through those who only use AI to save time. That's exactly what you need if you want your students to learn.
We already know the solution. What did colleges have students doing to ensure they knew how to use microsoft office? Three 50min lectures and two in-person 1.5hr labs a week for 2 credits, all dedicated to using word and excel. For a normal person in the 2000s, this is ridiculous to force on all students. It was not ridiculous 10 years before. Today, students again need the forced in-person learning for office applications.
For CS, homework should be done how you want. Homework is practice. Let AI assist you, but you better learn it because you'll be coding some part of it in lab with a physical programming book as a reference. They can lock down computers for labs to prevent web access or limit only to certain websites. Since labs are weekly, it will be obvious if someone is not learning as they do the homework. It won't be an F on the final as the first sign of trouble.
Schools can do things to improve student resources. Self host llms that are not corporate controlled that every student can use. The school can provide a search engine that is useful and not google's marketing. You can offer an education proxy that caches the pages you look at when doing education work, preserving anything you cite. This can all feed into the school llms for rebuilding to keep them trained on current info.
There are no projects anymore that are viable for a sophomore to complete that are difficult for modern AI tools.
I had many classes in college that didn't administer homework or take-home assignments. All we had were just exams. Most of those professors operated on the logic that if you wanted to pass their class, then you would do the (suggested) work.
Also, I never had a computer science class where exams were administered on a computer. I only graduated like 10 years ago, too. I shit you not, even my x86 assembly class was all pen and paper too, and yes, we were graded for code accuracy.
The AI stuff aside, this comes across as a professor who feels threatened when his course isn't taken as doctrine. Being called a cheater for using external resources is ridiculous to me.
Yeah, and sizeof and malloc are very easily things students could have been exposed to in high school or their own studies. I’ve seen C books that use them in the first chapter.
It’s fine to say solve this problem using the following tools, but not great to say “if you find another approach, I assume cheating.”
It did not say "I assume cheating". It said you're probably cheating, as in of you look at the population of students who are using such constructs, over 50% of them are using them because they are cheating. Given that accusations of cheating were not made when such constructs were found, it seems like that probability was not factored in when deciding whether or not to advise someone.
It does not say that one cannot use those things, it says that one is likely making their life more difficult. Implementations that utilize those things are, with high probability, more convoluted and involved. This slide is specifically around the time we introduced fscanf() and friends. If one is using the things mentioned in the slide, they're likely partially re-implementing *scanf() on their own. Which, while certainly a good exercise, misses the goal of developing familiarity and understanding surrounding *scanf().
While there are certainly a handful of experienced students that use these things, ten years+ of experience shows that most are copying from elsewhere. A simple, brief meeting with the student is easy enough to ascertain that, however. And that was certainly a part of the process.
Regardless, that alone would not have triggered an academic integrity investigation and, again, there was no prohibition here.
FWIW I really was happy to see your whole post.
As a student, _especially_ in the weedout classes where your performance in those classes decided if you could get into that major (I went to a school where you applied to majors 1-2 years after starting) it was incredibly frustrating to see other students cheating. When applying to a major that was well past capacity and very competitive to get into, it very much felt like a 0-sum scenario where cheating students were screwing the rest of us over. How awful would it have been to be declined from your major and have your career path severely impacted because so many students were cheating with 0 consequences
It was even worse when professors were notified but couldn't do anything about it (for example, a student next to me in an exam kept pulling out his phone but hiding it when the professor was nearby. I even notified her during the exam but she wasn't able to catch him so as far as I know, it never went anywhere.)
Similarly, I was a grader for a 300 level math class where about 50% of the class had blatantly cheated on an assignement (the solution manual on chegg had mixed up the questions so anyone who actually read the problem would've caught it because it was the wrong problem). I notified the professor and he said he would mention it in class but no other consequences occurred - they didn't even get 0s on the assignment!
All of your thought processes were very reasonable and I thought the best slide was comparing exam scores by people using LLMs on assignments vs people not using LLMs on assignments. I truly think that will convince people who otherwise might be borderline failing and desperate to pass that using an LLM is not in their own best interest.
I'd be curious to see (both as myself, and I think it'd be very convincing as a student in your class) the overall class grades for people using LLMs vs not. I could see someone reasoning "ok well if I get 100% on the assignment and scrape by on the exams, that still gives me a better grade than doing badly on the assignments (without an LLM) and doing slightly better on the exam"
Edit: I always thought for some reason that cheater would get their karma in their own way but in the real world, there are no consequences. I'm plagued by ads for ai-assisted ways to cheat on interviews every time I login to Blind. It actively is making things worse for the rest of us
Maybe you would have. Maybe you wouldn't have. But they have to design these things for the 90% case. People using things that are far outside what's being taught—in the age of AI—are maybe more likely to be cheating than everyone going above and beyond.
Students are human. Many have been taught to focus on one end goal: a grade. They see a mythical program that knows 6000x more than is taught in this 200 level course; of course they want to use it!
This isn't the 90% case. At least half of CS students come in with at least a basic proficiency in programming. It's a big enough portion of the class that not boring them into ignoring the actual Computer Science parts while getting the other half up to speed is the defining issue of lower-division CS courses.
When I took my Freshman year undergrad programming course decades ago, I had already been programming in that language for at least three years. It would have been quite painful to have to deliberately not do things I already know how to do because "I didn't learn it yet."
This was definitely a problem in classes before AI. I had a number of infuriating teachers who insisted on failing me for doing X, because "you're going too fast and we haven't gotten to X technique yet."
Decent professors distinguish between "using stuff we haven't taught yet" and "trivializing the assignment by using stuff that wasn't intended". If the assignment in an Algorithms course is to implement a square root function, calling "sqrt(x)" is obviously inappropriate. Structuring their solution using a do while() loop rather than while() loop is not.
I remember being marked down for using a ternary in Java for an introductory programming course at uni because the teaching assistant didn’t know what it was. Pretty minor loss of marks but it always annoyed me.
That said, having to do the introduction to programming class at uni despite programming for like a decade in high school was the bigger annoyance.
Same. I taught myself C when I was in high school, back in the early 90's. Fortunately, in college, I was able to test out of the intro classes for credit.
My professor in a C course said anyone that uses `goto` gets failed. It was quite liberating for me to then end up writing `goto`s for a living
Are you so graybeard that structured programming was interesting enough to be part of the focus of a C course? Because I can see in that context requiring the use of appropriate control flow structures. Nowadays, I can't imagine in what context students would be so eager to reach for GOTO that they don't preferentially use other control structures.
This was in...2013? I think the main point was just to stop students from footgunning themselves
I'm surprised by the negative response towards this professor. The syllabus was extremely clear up front, and students did not respect it and decided to see what they could get away with instead. Yeah there's some weird rules to follow, but the homework seems fairly scoped to smaller problems and it's just one class out of a few dozen.
Because it doesn't matter if he is right or wrong. A bunch of kids failed an important class and he didn't do a damn thing to address it. The teacher has failed. If you were his employer, would you think his methods are effective with the current students? I don't. Kids are lacking so many experiences, its just not reasonable to expect self study on everything. It also isn't required for learning.
I don't get why this is so hard to address. Make the homework a small part of the final grade and use in-person labs to ensure students are learning how to do the homework. Take advantage of llms as a tutor and helper, it is silly not to use the tools we use for so many classes already because we want kids actually doing the work. No one needs to shine light through a prism or swing a pendulum to pass phyiscs, they have labs and still make you do it.
In a lab, you can prevent the use of an llm and control the websites accessed. You can even put a relevant programming book at each station. You can do individual work where people all work alone and don't talk or you can do group work and vary during the semester. Make it a 3hr lab twice a week if that is what it takes. Phones stay in bags or pockets with dnd set.
I also think schools need to provide in-house llms for students to use. Labs, homework, research, etc should all be done using llms hosted by the school. It's crazy if this isn't already happening at every major university. Why give all that student data to a corporation using it to make your students jobless? If you keep all the search and llm use in-house, you can use the data for education improvements and research open to students.
Anyone who is lame enough to cheat on an intro to C class is doomed anyway; the AI they cheat with will be the AI that replaces them as a worker because they didn't learn anything. Give 'em hell, Prof. Turkstra!
Professors will be the last to admit their skill set is outdated. Those folks and academia in general aren't outcome-driven. They don't care if students will be ever able to find a job or not. They just want to educate like they got used to, and want their money. Or is it me being jealous? I had to change. Change a lot, learn a lot in last couple of years. But what about those folks who still use Borland compiler? I can imagine they're comfortable in academia, and can be in the next 10 or even 20 years - depends on how aggressive they're defending their old-world principles.
> Second, while the percentage of students identified was substantial, it was still less than 50%.
No doubt the student body will learn this lesson, and in the future the number of students identified as having cheated will be near zero.
It is certainly lower this semester, but not near zero (it was never near zero, even before "AI").
I was an assistant teacher at a private engineering school, teaching a toy kernel project.
We had to add a new significant assignment which was the student sitting down next to me and we both look at their code and I ask questions.
Our policy for the project wasn't even anti AI, it was simply "do not submit code you do not understand" so obviously we had to test understanding, but already with 90 students it was hard, I don't know how you would manage something like this with 500
> This form must be completed .... Failure to complete the form will result in an unfavorable letter being sent to the Dean of Students along with further potential disciplinary action.
> Through careful analysis and manual review, we have identified what we consider to be clear and concrete indicators in one or more of your homework assignment solutions that it was partially or entirely generated by an AI/LLM tool like ChatGPT, Claude, Copilot, etc....
> You are required to submit a response to this form....Failure to respond will result in a grade of 'F' for the course.
> Honest responses that align with our analysis will result in a score of O for the identified assignment(s).
> You also have the option to meet and dispute our findings. Should you choose to do so and we determine that our original findings of AI/LLM usage stand, we will instead apply the standard course policy from the syllabus and assign a grade of F for the course.
The above letter sent to a bunch of students.
> It is an open question whether this is reflective of their actual behavior or instead something that underscores the purported, and unintentional, potentially coercive nature of the form.
Sounds like the professor discovered (or disputes the possibility of?) the concept of forced confessions.
"Honest responses that align with our analysis": did the professor consider the possibility of dishonest responses that align with the analysis?
"Should you choose to do so and we determine that our original findings of AI/LLM usage stand" Not much needs to be said about that.
If the professor was confident enough about the analysis to send the letter, then all of the students getting the letter should have gotten an F and no letter at all.
It's just a textbook, garden-variety power trip. It's neither the first nor last time this professor will go on one.
https://www.ratemyprofessors.com/professor/2231495
My “power trip professor” alarm went off reading:
> If we find reason to believe that a student or team has cheated on any assignment, we may inform the student or team promptly, or we may decide to silently accumulate evidence against the student or team on later assignments.
Why role play what you may or may not do if you encounter cheating.
The university has rules about how academic integrity is handled independent of the syllabus.
Including this line in the syllabus, and again at the top of the blog post is weird
That's not roleplaying, that's the basics of how you handle stuff like this. For lower-division level technical assignments, there's only so much free expression to extract signal from with regards to copying per assignment. Any individual assignment could likely be appealed as coincidence. If the same students turn in 2 more assignments that are suspiciously similar to the same other students, then you can have a pretty iron-clad argument for cheating.
Students cheating at lower-division university level courses is almost entirely a symptom of the students being poorly suited for the course material. Students who are interested and capable, but overwhelmed with the amount of work tend to just turn in incomplete assignments, or not turn in work at all. Encouraging the cheaters to drop while they can still get "dropped" rather than an "incomplete" or "F" is in everyone's best interest, as it provides an opportunity to consider a pivot to something that interests them more.
IMO, the overreaction by the professor was seeking further penalty for students who already dropped the course, or who would do so when confronted with their misbehavior. A student who cheats in CS is not necessarily a student who will cheat at chemistry, or business, or journalism. If they've already dropped out of CS, there's little benefit to be gained by pursuing the matter further.
> that's the basics of how you handle stuff like this
Anecdotal but not in my experience. The professor can inform students, collect evidence, etc without detailing each scenario in the syllabus.
Every CS class I’ve been a TA for had the same generic university provided statement on cheating.
The professors did not want to spend time thinking about or dealing with people cheating.
Every time MOSS flagged assignments they basically sat with me and the student and said “You are likely cheating. Please don’t do it again.”. No one ever admitted to cheating and no one ever got flagged twice.
edit: HN seems to loves people who steal, lie, and cheat. my bad
That’s an entirely different situation.
It would be more like a low level manager telling new employees what he would do if he ever catches them stealing office supplies.
The university has a policy on how to handle cheating. The professor doesn’t need to imagine potential scenarios in the syllabus
I recently completed my masters in CS. I agree with the sentiment of wanting to avoid LLM use in general, but that ship has sailed. No amount of policies and slides is going to be enough for this systemic issue. I remember all the policies and volumes of reading I would have to do to keep up with what was expected to be done in a very rigorous and usually pedantic way. Change to oral exams or something similar
It's beyond me how binary their thinking is when dealing with an issue as nuanced as AI. For instance euros have oral exams often and they are much more helpful on seeing if the kid has any idea on what they are doing. Furthermore it forces them to think through the problem more clearly since they have to explain it to someone else at some point (rather than some autograder).
It's part of the problem with academia being behind industry.
I had a similar problem when doing my MBA; Lean Startup had just been released and the whole "startup scene" was getting going. But I couldn't use any of that in my MBA entrepreneurial course, because no-one had written a paper on Lean. I had to write business plans (which, obviously, no-one was doing out in the real world). I was being taught a useless skill because academia couldn't adapt quick enough.
For these CS grads who are being punished so harshly for using AI, the same: the first thing they will meet when they start their new developer job is a Claude prompt. Their CS course will not have prepared them for this, and they will suck at using an LLM to generate code.
There is an argument that universities should not be vocational, and should teach people how to be better humans, not better employees. I agree completely with this, but that also means universities need to stop advertising how getting a degree leads to more lucrative employment. Which is it?
Because if it is vocational, and if a CS degree is supposed to lead on to a career in software development, then they need to teach orchestration not C. We don't teach prospective developers how to write machine code on punched cards any more, for exactly the same reasons.
> For these CS grads who are being punished so harshly for using AI, the same: the first thing they will meet when they start their new developer job is a Claude prompt. Their CS course will not have prepared them for this, and they will suck at using an LLM to generate code.
You're assuming that this policy continues for higher level courses. It does not. Use of LLMs is discussed, and in some cases even encouraged, in higher level courses like Software Engineering I.
I would argue that if they use an LLM at the stage of a lower-level fundamentals course, they'll suck even more at using an LLM for more interesting tasks. They also won't have the skills necessary to address problems when the LLM invariably doesn't do exactly what is desired.
That's good to know. I'd assumed that this was one of those courses.
> I would argue that if they use an LLM at the stage of a lower-level fundamentals course, they'll suck even more at using an LLM for more interesting tasks. They also won't have the skills necessary to address problems when the LLM invariably doesn't do exactly what is desired.
That's how it works now, kinda, with all us seniors controlling the LLMs. But we're learning a whole new set of skills that are only tangentially related to the skills required to actually write code. I can easily see that in a couple of years students are taught those new skills directly.
> I can easily see that in a couple of years students are taught those new skills directly.
Definitely. We're already getting ready to prototype a course that does just that.
> There is an argument that universities should not be vocational, and should teach people how to be better humans, not better employees. I agree completely with this, but that also means universities need to stop advertising how getting a degree leads to more lucrative employment. Which is it?
I read a reply to a similar question in a different HN thread a while back that stuck with me. It went something like, "The point of university is to get an education. The point of student loans is to get a job."
haha yes, agreed.
One of my hopes is that AI screws up the whole career thing so badly that we go back to education being non-vocational (because there's no point, everything they teach will be obsolete before the students graduate).
> Every student is also responsible for protecting his/her own work. If you inadvertently allow your work to be accessed, you may still be held responsible for facilitating an academic integrity violation.
This is insane. Absolute power tripping. And I say that being in favor of banning AI throughout the course.
I suppose it's to close a loophole where students can collaborate/cheat and then, if caught, claim that someone else copied off of them.
I'm a fan of zero tolerance policies for academic dishonesty. The dean getting involved and mitigating consequences for the ~~customers~~ students is classic.
The real crime here is those ugly slides
Only rivaled by the ugly website :-)
> Do not look at other students’ homework files and do not work with other students. At all.
This is a horrific policy. We work better with other people.
When I studied computer science, other students came to me for help. I helped them. We discussed problems together. I saw their code. They saw mine. No one copied: we shared approaches and style and thinking. I shared how I solved things. They learned my thinking. By explaining, I understood what I knew better. Sometimes I learned from them: an idea I wouldn’t have had alone, a new direction. All of us were better off. This was across multiple people.
It was in no way discouraged and so long as you wrote your work — while asking others for help was fine — you were okay.
Humans learn through teaching. We learn through discussion. We learn through interaction. Solitude — ‘do not work with other students. At all’ — is the exact opposite of a healthy learning environment.
I don’t care what their artificial intelligence policy is. I cannot state strongly enough how much I revolt against their human intelligence policy.
As for the lecturer’s warnings in slides: that reads like a trap.
The course Programming in C, and other courses like Compilers, have this policy across many CS universities for many years now. It s a strict policy but it has its reasons. Assignments on early semesters are simple and short, and if students cooperate on the implementation, the result would be almost identical for everyone, while very few would have truly thought it out by themselves. Most students would not go easily through the process to struggle in order to think about the solution, if they could just copy the answer from someone else.
Thanks for the reply. I get that, and maybe I am approaching it from the perspective of someone who wanted to learn (my friends too), not someone who just wanted to pass.
But the situation I described included the basics. I specifically remember explaining pointers to a friend and working through a simple assignment on C pointers with them, for example. Absolute basics, yet collaborative working helped them greatly.
Perhaps your policy puts TAs there in that situation, and I understand that.
This was the case at UofM. When it's time to collaborate, it was group projects that were less specified.
It's important to note that the students are encouraged to interact with the many teaching assistants that we have available. And, while that slide is somewhat harsh in how it states it, the larger lecture discussion surrounding it makes it clear that students are permitted to discuss assignments in a general way. When you get into implementation details, that's where the academic integrity problem arises.
For people that haven't spent time teaching lower-level programming courses, policies like this may seem overly harsh. But, remember, they're not solving particularly advanced problems at this stage. Instead they're mastering fundamentals. An analogy I like to use is learning to play the piano. If someone else is doing it for you, you're not going to develop the skill.
The main reason for the policy is that many students are not able to walk the line between leading someone to a solution and giving them a solution. The teaching staff know how to do this. When students help each other in situations like this, many times they end up just providing code or allowing someone to look at their code (ultimately copying it), etc.
I appreciate your reply, and thankyou. It did strike me as harsh, yes, and I was rather horrified when I wrote the parent comment. But I think I was also responding to the tone of the article, which struck me as almost setting traps - that kind of emphasis on the student being responsible for understanding something broad yet not saying what it is specifically in the broad material they need to look out for, for example.
Yet I also understand your view -- students providing and copying code yet not learning and doing work themselves. I completely get that needs to be found. Yet, reading the article I felt truly scared for the students who might get caught up in this yet I don't think would have actually done anything wrong.
> Yet I also understand your view -- students providing and copying code yet not learning and doing work themselves. I completely get that needs to be found. Yet, reading the article I felt truly scared for the students who might get caught up in this yet I don't think would have actually done anything wrong.
I get that, and I think that is partly what drove a lot of the initial turmoil when this started. But, the important thing to recognize (and this could have been better emphasized for sure) is that it's really easy to tell whether someone caught up in this is cheating. A 5 minute discussion is all it takes. And, yeah, maybe some could BS their way through a discussion like that if they had a heads up, but at that point 1) they've now learned at least some of the material and 2) you're not going to catch everyone anyway. Processes like this are somewhat like a police officer catching people speeding. The enforcement alone discourages the behavior, to a degree, and you're never going to catch everyone.
I agree. Here at U. Lisbon, we are following the recommendation by the recent MIT report: Evaluate students by talking with them, making it a social interaction, which they are lacking.
We need students to work together and share their experiences. Then we need each one to achieve the learning goals (and we should not assess work they did far from our eyes. We should assess their understanding and skills in front of us).
I had a CS professor who worked like this. We had assignments and tests, and they did have grades, but your overall grade in the course ultimately boiled down to convincing him that you understood the course material. As someone who excelled at the material, but could be a bit hit-or-miss on bothering to complete assignments, I loved it. Students who were used to optimizing their efforts and obsessing over each point hated it. It didn't come up in that course, but the issue with such a grading scheme is that sometimes students and professors really don't see eye-to-eye. Most of the time it's the professor's fault, holding some sort of prejudice. A grading scheme like I described would be disastrous for such students.
This is a good approach, and is used in some situations. But, it has challenges in terms of scalability. It is not impossible for larger enrollment courses (this course had 600 students), but there are certainly trade-offs in terms of consistency in grading, overall student experience, etc.
Sounds like to me maybe universities should stop focusing on being cash grabbing machines and instead properly teach with a useful instructor-to-student ratio?
Sorry, if your answer as to why using a beneficial pedagogical approach is your class sizes are too large, then it sounds like your class sizes are too large!
You're not wrong. But, that type of concern usually needs to be addressed by someone at a higher pay grade than professor ;-)
learning to think is orthogonal to american universities, who's goal is to get government money via fed backed loans to fund professors real hobby: doin cool science
standardized and written tests, online school, homework, etc is just a way for professors (and increasingly, just TA's) who have no interest in teaching, to get the job done with less time and more efficiently.
Oral exams solve this problem entirely
Some of us are perfectly happy to learn in silence and solitude, and we thoroughly excel at doing so. Sorry that you don't. The broad sweeping "we humans" language gets a groan out of me. It's the sort of mentality that leads to unproductive gatherings in the working world where people want to force "learning through collisions/synergy".
As someone who's graded lots of undergrads I can tell you that cheating is prolific. It's easy to identify who is copying code, but it is hard to prove. My wife teaches in another department and says the same thing. AI use has grown exponentially and it's really destructive to the education process. Though I have more complex thoughts on this and that it is just exacerbating larger problems that no one wants to actually solve because they'd be very hard to solve in a way you can have clear metrics for the bureaucrats (including industry). Proper evaluation is unfortunately always going to be fuzzy. Metrics are in contention with making a meritocracy, especially if you don't have a deep understanding of said metrics
naive to think people would follow this policy in the spirit in which it’s intended, but even then it clearly selects for the most social animals amongst your students instead of the most talented. And discussion alone is all it takes to describe the right approach, so if that’s the test, it’s meaningless to protect the mechanical syntax of implementation.
Clearly in this world the frats and the sorority members get the highest scores, and after that the best looking, the most amusing or charming etc.. so unless that’s the intent the whole approach is perhaps worth reconsidering.
If you want to teach to improve your understanding the normal thing is you become a TA so that access is open.. not open every assignment to ad hoc collaboration so you disadvantage half of the students on every assignment
This really reads like some sort of incel manifesto. The odds aren’t stacked against you nearly as much as you make out lol.
It may be surprising to you but, caring about fairness and democratizing access to helpful resources doesn’t actually require any kind of disadvantage or a complex. What’s your deal though, since you want to attack that?
I'll pile onto the agreement stack. Abso-100%-freaking-lutely. I worked with my friends through all our assignments, bouncing ideas off each other, talking through confusing bits, etc., and then doing the work ourselves. We didn't ask "how do you write this line of code", but "how are you approaching this problem?". I learned as much from the discussion as I did from the implementation. Throwing that out is absolute madness. No way.
> You may discuss assignments in a general way with other students, but you may not consult anyone else's work.
So... what it says?
It wasn’t general at all. More like, “I’m going spinlock poll this memory mapped address until it gets the trigger value, then…”
I sat on a university student disciplinary panel. I assure you that your experiences are biased. There are plenty of people that flat-out copy each other.
GP was describing the platonic ideal of why we have universities instead of just textbooks, so I'd venture a guess that you're the one with a biased perspective here. Sure, some students cheat, and the appropriate penalty is throwing them out of the program. But pursuing that end to the point that you forbid students from engaging in normal collaborative learning is nuts.
As you point out, there are plenty of students who flat-out copy each other. Filtering out those students does plenty to enrich the remaining students, you don't need to go on witch hunts to find ever-more subtle cheating.
Given the professor's professional interest in cheating-catching tools, I'm suspecting wanting to get good, interesting data motivated a lot of their behavior here, even in ways he might not consciously realize.
Any policy that asks students not to use AI and trusts that they will comply is giving a massive advantage to dishonest students and punishing honest ones.
This article sounds like the professor was reprimanded by higher powers and tried to dig themselves out of a hole for having a badly structured grading system.
Students should be encouraged to use AI in their learning, because they will anyway. The understanding should be that they will be assessed in supervised environments so they need to learn the work either way. This gives the students incentive to learn without fear of using the best tools available to do so, and provides evidence to the powers that they actually did so.
If you remove the mention of AI from this post, the authors process and setup of the notice to students is absolutely mental. The intentions might be good, but damn if this isn’t a comically inept way to set up a kafkaesque nightmare rather than a place for actual learning. The defensive attitude doesn’t help either. Just the sheer energy that went into all the steps mentioned here, not invested in any even superficial attempt to perhaps adapt. That isn’t to “give into AI” or cheating, but if 50% of students break your rules, more rules and litigation about the rules are not the next step to take.
Using LLMs for the classroom assignment is like using a forklift to lift weights in a gym: completely inappropriate and atrophic.
That some 50% are using the LLMs rather than spend effort learning is not a cause to remove the ban of LLMs in the classroom.
90% of the power of an LLM is being yourself competent enough to judge its outputs. Besides, do we really need more humans coding like LLMs?
My comment explicitly did not focus in the use of LLMs and in fact much of the post doesn’t. Leading with the depressing “work alone” slide, it focused on similarity, not AI use. Between students, ruling even “inadvertent” exposure as the same punishable (and literally literally undeniable!) offence. There’s millions of ways to structure a class for learning, this myopic focus is not one of them. Ironically the setup ignores the very basics of human collaboration. AI merely amplifies extremes, it’s a factor but neither the cause nor culprit here.
The title includes “retrospective” but the learning and changes following such a process are absent.
> Leading with the depressing “work alone” slide
working in a team is valuable. but that value is proportional to the teammates chosen. teammates who have worked alone and can think through their decisions are more valuable, thus it is this way.
>AI merely amplifies extremes, it’s a factor but neither the cause nor culprit here.
you may have missed this part: ~As discussed in the paper, we have also run the tool against earlier semesters with very compelling and interesting results. In particular, the indicators that we use are virtually nonexistent prior to 2024 and their presence increases rapidly over the following years.~
So it stands to reason that LLM's are in fact, the culprit here.
All these analogies are kind of tortured as they ignore the elephant in the room. The students want to be there to get their job getting paper. They aren’t there to learn.
If society offered some reward for lifting the weight while not enforcing how, forklifts would absolutely be used to lift weights.
The job-getting paper ceases to be meaningful if it no longer verifies competence. If it merely verifies the ability to prompt an LLM then nobody will pay a premium for people have it, and then nobody will pay to obtain one.
... Thus the enforcement?
Yea, this letter could have been written about garden variety "suspected cheating" 20 years ago. And the professor's attitude is the same as a couple of professors I remember back then:
Prof: "I know you cheated."
Student: "How do you know it, can you prove it?"
Prof: "I have ways, trust me I know."
Student: "I know the material, I'll re-take the test right in front of you if you want."
Prof: "Either admit that you cheated and get a F, or don't admit and I'll send you to the school's academic integrity department."
This is the professor who had beef with a /middle schooler/ for copying some content on how to use a calculator from his website. His defensive attitude is clearly indicative of the massive, massive chip on his shoulder. Just reading this triggered me to my experience with similar professors in college about a decade ago. What a jackass
> Additionally, the kid that I "stuck it to" wasn't even in high school - I'd guess middle school or possibly even grade school.
"Okay, but that's worse. You do get how that's worse, right?"
To be clear to anyone reading this, that is a verbatim quote from the man's own FAQ. Wild.
It's worth reading the whole FAQ.
It was an AP History assignment he did with friends at school. Not content on how to use a calculator.
https://turkeyland.net/projects/atomic/atom.html
It was also back in 2001, when he was 18 years old. The dangers of having a site spanning multiple decades. Heh.
The insistence that you must "work alone" and that similarity between submissions will get you flagged would make me really uncomfortable. I think I would be afraid to talk to the TAs lest I get poisoned by their knowledge.
I think there are two somewhat orthogonal things addressed here. The first is adapting in the face of AI, and that is happening albeit not instantaneously. The other part is simply ensuring students are working within established policies. In some sense at this level using AI is similar to handing a student a calculator while teaching them to do basic mathematics. They're not going to learn addition and subtraction if they don't actually do the work.
As far as the "comically inept" setup, I think that's acknowledged in the retrospective - including an apology.
Can't they just do in-person testing? Build a cluster of cheap, crappy dumb terminal from 40 years ago. Have students type out the source code live. Grade the work. Hell, let them use reference books! Just don't allow re-typing AI-generated code on notebook paper. They can cheat all they want in the interim, it'll all become clear on the midterm and final exam.
Haha calling the fire marshal ahead of time "to ensure we remained within the capacity constraints for the facility" is an amazing touch!
This sounds snarky, but I don't quite get why. That's exactly what was done, and in the end nobody needed to be (nor was) removed from the building. Although those not enrolled in the course were relegated to the surrounding open spaces/hallway. Which was a good thing, because lecture attendance on that day was near 100% (i.e., almost all seats were occupied by students).
it's hard to rein in cheating with LLMs in the classroom. this guy is trying at great cost to his personal life - i mean, if not in academia, where else are mean people supposed to achieve great things?
imo watermarking will do little to reduce this awkward, personal, adversarial thing. as always, technology cannot 100% solve, or even 50% solve, a political problem.
> the decision that was reached was to reverse the retroactive use of the tool on earlier homework submissions... Students that had dropped the course would be permitted to re-enroll.
it's too bad the professor completely capitulated. like people have so much angst against H1Bs, so much excitement about unionizing, agitation about layoffs or whatever... look, 40% of your classmates are cheating. 40% of your colleagues are cheats. the call is coming from inside the house!