Cheating was widespread even before LLMs! I can’t imagine what it’s like now!
Who is actually surprised that Economists cheat using LLMs? They are in the same cheater class as lawyers, who have been caught cheating in court using LLMs.
I’ve grappled with this before, and even before people were using AI, take home exams were always open-book by default because of course people will look things up. The irony is, since AI, I’ve had a lot worse essay submissions-- they simply get shit wrong a lot more relying on AI (especially in my statistics class).
And my stats class only needs to do basic stuff to show work; I let them use excel/sheets to check work. They can’t even type in “=correl()” properly and would rather take a photo of the data and send it to ChatGPT. AI gives them wildly wrong steps and it’s so obviously wrong…
Midterms were take home, honorable closed-book; final was in person.
Of the 89 students who did the midterm exam, only 59 showed up for the final one. And of the 27 who did not show up, 22 had scored a perfect 100 in the midterm exam.
It seems to me that it’s important for those 22 to be expelled permanently for academic dishonesty. The president of Harvard was kicked out of the position when it came to light that she had “plagiarized” herself. This is far more egregious.
Serrano did not void the midterm exam, but warned students that the final one, which counted for 50% of the final grade, would be held in-person. He also said that if the grade distribution was not similar to the midterm, only the final exam would be taken into account. The average score dropped to 48 out of 100. Of the 89 students who did the midterm exam, only 59 showed up for the final one. And of the 27 who did not show up, 22 had scored a perfect 100 in the midterm exam.
“The empirical evidence of fraud is overwhelming,” says the professor, who has decided to make changes for the coming academic year. First, the weekly exercises will not count towards the final grade, as these could be done with AI. Second, no more take-home exams, no matter how appropriate they would be.
Bruh…
Long before chatbots people were 100% cheating on “closed book, take him tests” there is no fucking way they weren’t.
If you had told a class 30 years ago the same thing, the same result would have happened.
Saying if you don’t match you only get the worse score, is going to discourage a shit ton of students.
The real problem is “closed book take home”. And professors expecting students to not open the book. It’s nothing about them, it’s the current culture. If you don’t cheat you can’t compete. Elite places like Ivy League schools will always be full of cheaters in this climate, because they’ll out compete people with ethics.
Seems like the solution is: don’t do exams this way
This year, the economist decided that both the midterm and the final exams for his course would be of the take-home, closed-book type (there is a certain tradition of this at Ivy League schools). “It’s a very nice kind of exam, because as you’re giving students practically unlimited time to complete it, it lets you make it harder than normal, to see how far they can go.” In this case, Serrano changed some of the model assumptions they had seen in class, and asked students to demonstrate whether certain statements were true or false under the new assumptions.
It’s funny to watch people catch up with what’s been happening for years and what was foreseeable at least since 2020. The moment I saw an early version of chatGPT in 2019 I knew pretty much the whole story, although I didn’t account for the massive infrastructure build out backlash because I seriously underestimated the resource consumption.
