Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

CheatGPT

blog.humphd.org

Post reply on HN
Print view View original

111–120 of 544 posts

Re: CheatGPT

#111

Post by hgomersall »

kerpotgh wrote: post #96
ttul wrote: post #90

The best counter to ChatGPT cheating that I’ve heard of is to get students to orally defend their assignments in class. No computer is going to come to your rescue when you have to explain why you wrote the code you wrote. Encourage the use of ChatGPT and other tools. Put the emphasis on understanding what the code does and why. The tool may help to explain this to the student, but if so, that’s fantastic. No need to…

There isn’t any time for this in a 500 person public university classroom.

When I was at uni, there was no marking of assignments. The "examples papers" were basically your opportunity to understand the course before a supervision when you could discuss them. If you didn't do the examples, the supervisor didn't really care (typically a grad student that had enough of their own worries). My point is, if I'd cheated, it would have ultimately been my problem when it came to sitting the exam.

Re: CheatGPT

#112

Post by corbulo »

dcow wrote: post #81
corbulo wrote: post #79

Earlier quoted context omitted.

It will be a hard sell since it or similar tech has already been long embraced commercially. A bit like forcing students to use an abacus while calculators are around. The student should still be responsible for what they submit and all of its errors of course. But what should really change is our methods of assessing students. The University model for undergrads has already been long broken in need of fixing. I don'…

> A bit like forcing students to use an abacus while calculators are around. I said allow students to use ChatGPT. Just make it clear that pasting answers from it verbatim, just like doing so from SO, is called plagiarism and does not benefit the academic community or themselves. There will always be cheats. Agree about shifting evaluation methods.

What about when it gives the 'best' word choice? Should a student change it arbitrarily?

Re: CheatGPT

#113

Post by Thorentis »

stillsleepy wrote: post #91
novok wrote: post #53

The failure mode of things like ChatGPT is it can make wrong answers confidently, subtly, and if you don't have the skill to audit what is wrong with the answer, then it can be fairly catastrophic. So instead of making questions generative, you make them audits / debugging type ones. Use ChatGPT to generate a result after several iterations that is wrong and then ask them what is wrong with the result. Since ChatGPT…

the other week i gave chatgpt a simple multiplication problem that it got wrong. very simple problem like 86 * 0.0007 or something. but ive been working with chatgpt for 4-5 weeks now and that wrong answer doesnt make up for all the "good answers" that are usually not perfect. like one day i needed to COALESCE in mysql. i didnt know that, but chatgpt did. theres a few times i would have written a function the complic…

Multiplication problems are not language problems. There is no data in the training set where there is a likely probability of the next token in the "86 * 0.0007 =" sentence being correct.

People need to stop treating ChatGPT as a computation engine. It is not wolfram alpha. It is not google. It is fancy autocomplete trained on a large subset of the internet.

Re: CheatGPT

#114

Post by _cenw »

singularity2001 wrote: post #107

Maybe take the approach many physics classes embrace (in germany) : All tools are allowed in tests: calculators, books, as long as the correct answer is derived with "correct derivation path". If chatgpt makes some problems too easy maybe it's a good thing because we can raise the bar. Find problems that require true understanding beyond auto complete / copy pasting.

GitHub Copilot quite often delivers comments alongside the code for a prompt.

Re: CheatGPT

#115

Post by singularity2001 »

vsareto wrote: post #106
tommica wrote: post #102

Earlier quoted context omitted.

Yeah, this is a good one - does it matter if you wrote the code, or a computer did? If you can explain how it works, and why it is like that that should mean that you understand it and have learned the important parts.

You could ask it to explain line-by-line for you and memorize that enough to give an oral presentation. You may not get the best grade, but it might be enough for a passing grade.

If you memorize it that's half way to understanding, so fair use

Re: CheatGPT

#116

Post by Thorentis »

jimmar wrote: post #104

I'm a college professor. I'm requiring my students to use ChatGPT to aid them on some assignments this semester. Results are mixed so far. I agree with the author, "One of the things I'm thinking about is that I might need to start teaching students how to use GPT/Codex and other LLMs in their [assignments]." One of my top students submitted a sub-par assignment because he relied too much on ChatGPT and provided litt…

> I'm requiring my students to use ChatGPT

God save us from this horrible future.

Re: CheatGPT

#117

Post by neilv »

I see a ton of people trying to justify their use of Copilot/ChatGPT (or rushing a startup to cash in on LLMs).

Maybe that conflict of interest is why there's very little talk of it being based on plagiarizing and license violation of open source code on which the model was trained.

We just suffered through a couple decades of almost every company in our field selling out users' privacy, one way or another. And years of shamelessly obvious crypto/blockchain scams. So I guess it'd be surprising if our field didn't greedily snap up the next unethical opportunity.

Re: CheatGPT

#118

Post by otterpro »

If I were a teacher, I'd give written exams for coding, as we used to have back in the 80's when we didn't have enough dedicated computers for everyone in the class, and most of kids didn't have a home computer to work with. So instead, we'd implement simple algorithms on paper for in-class quiz/tests/homeworks. I remember in high school, we had an exam, and one of the question was to write a for loop counting from 1 to 40. It might be harder to grade, but teacher could usually tell if the code looked good or not. On one occassion, I thought I had written the code correctly but teacher didn't, so I had to demonstrate the code on the computer for the code that I had written in the test and show the teacher that the code ran as expected.

It's kinda like whiteboard coding, but it's a lot less stressful since the test is mostly written on paper without the scrutiny of an interviewer. Obviously, we can't create a react web app, but at least students can demonstrate fundamentals, which are what we should be teaching in the first place.

Another solution would be to create a new programming language that doesn't exist yet for the students to write code for their assignments or perhaps use an obscure language that AI/chatgpt isn't aware of... perhaps we can go back to Ada or Modula 3.

Re: CheatGPT

#119

Post by humanizersequel »

rpearl wrote: post #88
hutzlibu wrote: post #38

Earlier quoted context omitted.

If the point of school work is to prepare for life - then you should mainly learn, how to get a job done. And if they want to teach special skills, like writing essays without computerhelp - then you can test that onsite.

The point of writing an essay is not to learn how to produce an essay. It's to learn analytical thinking, research, and argument skills.

Knowing how to produce an essay is exactly the same as "analytical thinking, research, and argument skills" with the added challenge of making it legible to a reader — which is what makes those skills useful.

Re: CheatGPT

#120

Post by kube-system »

wpietri wrote: post #110

This is definitely going to require changes in interview practices for a lot of places. For those looking for an alternative, may I suggest pair programming? I've been doing pairing interviews for years. These days I have a standardized, practical problem, something that's reasonably like the work. (E.g., let's use APIs A and B to build features X, Y, and Z.) I let them pick their preferred language and tooling, so t…

The emergence of this problem is exactly why I don’t like giving people programming exercises that are overly explicit.

Most coding tests just tell people what to write and then have them write it. Real world problems are more complicated. Instead, tell your candidates what your problem is and then ask them for a solution. Let them write their own requirements. It’s a lot harder for language models (and developers with poor problem solving skills) to solve these kinds of questions as well.

Post reply on HN