University of Waterloo withholds coding contest results over suspected AI use
thelogic.co
thelogic.co
The first stage of the CCC is written in pretty heterogeneous conditions. There are several thousand competitors across the world. Students use a language of their choice and their work is graded by their teacher against a rubric sent out by the contest organisers. I recall that the teacher has to send the work to Waterloo, postmarked within a couple of days after the contest.
Sometime in the summer, the University of Waterloo then flies out and puts up the top 15 or 20 students in that more informal contest to compete in what was the "CCC Stage 2" and is now the "Canadian Computing Olympiad." This second contest is done under more controlled conditions and its results largely determine which four students will represent Canada at the International Olympiad in Informatics.
The first stage has always relied on teachers' honesty. The contest organisers mail contest packets to teachers in advance of the contest. Teachers have some time between contest day and when the results need to be mailed back to Waterloo. Students are comparing notes and discussing the contest immediately after the contest.
Nonetheless, in my experience (more than 15 years ago), all or almost all of the students who make it to the second stage deserved to be there. I hope that continues to be true.
[1]: https://www.washingtonpost.com/technology/2023/08/14/prove-f... (archived/paywall bypass) https://archive.is/BoVlG
The first stage (this one, with ballpark 10000 competitors) is distributed and done in very heterogeneous environments, and it relies on local high school teachers taking time out of their day to set up whatever's needed, proctor the contest, and grade the results.
"could have", yes, but mostly didn't until LLMs showed up.
The problem is that LLMs lower the bar to cheating: using Stack Overflow still requires a pretty solid knowledge of the domain (and in a professional coding context, usually wouldn't even count as cheating). The LLMs also let you do this much faster.
I don’t know if I would consider this modern coding using modern coding tool chains anymore though. It might still be a nice competition to run, but it is increasingly displaced from what it means to be a professional programmer (I’ve stopped caring about these competition wins on resumes a long time ago).
Some of the best engineers (I mean that in the broad sense, not just “were good at coding up algorithms”) I had the privilege of working alongside with scored highly on these competitions.
Outside of being highly ironic, this is how you get a GAN arms race :)
... I guess so, but how it is fault of someone who did not cheat? They are lumped in together, like it's their fault too. Not that I don't get that, but still.
E.g. if half the students' submissions had an 80% chance of being AI-generated then something is very fishy in the population - perhaps a 99% chance a quarter of the students cheated - but 80% is not really high enough to punish an individual, especially since AI detectors aren't very reliable.
It's also possible to severely lock down the network connectivity of the computers so they cannot access the internet at large but only the contest servers during the competition if there is an online submission required.
There's no solution at this point for this year.
Return to tradition[1].
We had a couple of people shrug it off and do it in notepad and csc quite happily. They were of course hired.
It was referred to as "book code."
I love this comment because in 20 years coding with no AI will be "coding like it's 2020 again".
The challenge is figuring out how to provide a novice (high school students) with a skill appropriate challenge to assess them on that is also not trivially solved by the use of an LLM, and you likely can't. Or you go the IT route and restrict what's available on and from the computers (see my comment about how ACM and IEEE contests used to do things, they may still run that way but I'm uninvolved so don't know).
I have done this competition (I think) and what your suggesting just is not reasonable. This is a distributed competition administered locally on a volunteer basis. This isn't some formalized test & trying to get Waterloo to pay for all the local proctors would result in this competition shutting down which would be net not good. I don't have insight into how the competition has changed in the 20 years since I took it so I don't know why they don't have local proctors or lost faith in them. Maybe COVID-era adaptations?
Honestly, I think the better solution is to adjust the difficulty to be AI-proof but it may be hard to do so at this level of academia.
Putting that aside, you don't just sue for money, you can sue for different outcomes and injunctive relief. For example you could sue to have the results released as planned.
Injunctive relief for not publishing known to be incorrect test rankings? Not likely.
A FOI request might get you access to the rankings if you really cared... I'm not quite sure what the contours of FIPPA (roughly equivalent to America's FOIA) is with regards to our universities but I believe it generally applies to them. At the same time it could easily fall within privacy limitations.
Right now educational institutions are facing budget shortfalls and responding by expanding remote education opportunities while compromising the quality of education not just because of cheating but because often the content and social circumstances are inferior. We are making short term decisions as an easy-fix to budget issues that will have substantial long-term consequences.
The result is that the value of an American college degree is plummeting (ironically while costs still are increasing). This is part of the reason GenZ are starting to revolt against various institutions - they are the ones being screwed the most.
The right path forward is acknowledging this insanity in the budgeting process, employment process, and loan markets. We won't do that though because it means admitting how much has been fucked up - institutions will double down and triple down in the face of DOGE threats.
This Canadian university run high school coding contest isn't part of an American college degree. In fact it has no relation to any of that description. It isn't American. It isn't relating to a college at all by the Canadian understanding of that word (we distinguish between college and university, and Waterloo is the latter). It isn't for university (or college) students or related to a university (or college) degree.
> Look at medical schools, we're not pumping out chatgpt doctors because we have actual in-person test proctoring that includes checking biometrics, taking photographs of the individual for posterity, oral examinations, etc.
Extracurricular contests for high school students do not have the same funding or standards as medical exams to become a doctor for obvious reasons (in Canada, where I participated in these, but presumably anywhere).
Yep. Related, but the software engineering industry did actually have (for a brief time) a PE Exam but it was sadly discontinued.
https://ncees.org/ncees-discontinuing-pe-software-engineerin...
There's still a PE Electrical and Computer: Computer Engineering one but its definitely more focused on hardware.
Though most software developers actually go through CS not Software Engineering and that doesn't get you the PE designation.
https://www.peo.on.ca/apply/become-professional-engineer/tec...
Medical schools don’t take in 18 year olds in North America. They can have more in depth entrance exams since the people come in are more mature.
> This is the reason GenZ
Or perhaps it may be the misinformation and rhetoric being used to inflame tried and true anti-intellectual sentiments.
Why do you need to be more mature to take an exam? We're not talking about toddlers - many (most?) of them are legal adults.
All of which is beside the point since this conversation is focusing on the most prestigious universities around. Most universities and colleges in North America don’t need such exams since they already accept basically anyone who applies.
I truly do not understand your point at all.
> proctoring that includes checking biometrics, taking photographs of the individual for posterity, oral examinations, etc.
While I personally do actually believe test like that would be better than the kinds of exams being taken by students while at university I don’t see it as a realistic option for entrance exams. That sort of thing is simply not part of North American educational culture to a large degree. And for medical schools people practice and study and drill for months in order to prepare themselves for it.
Also, for most colleges and universities it simply makes no sense to introduce such an exam (or any at all) since they accept everyone who applies; they need to in order to remain financially viable.
This isn't even an entrance exam. It's an extracurricular event for high school students that people primarily write for fun...
> since they accept everyone who applies;
This on the other hand doesn't describe Waterloo CS at all.
A) I know this. B) If it gives you a significant benefit to admissions, it will quickly become a metric to be gamed.
> doesn’t describe Waterloo CS
Sure and I’ve said as much in another comment, but the person I replied to was making a broad comment about higher education in general.