since there has recently been a lot of cheaters and AI users why can't we report them like having a system for reporting people with suspicious code and if someone gets more than a threshold of reports their account would get banned or get flagged or something, I think this would be better than having to wait for the members of the Codeforces team to continuously check for cheaters or have to depend on those AI detects as I feel like they don't work 100% This is definitely not a perfect solution but it might help a bit
Hey Marzouk, your idea of a community-driven reporting system is quite innovative — it could act as a "crowdsourced" layer to flag suspicious activity more quickly.
However, to avoid abuse (such as false reports targeting strong coders), the system will need safeguards, including but not limited to:
(1)Thresholds: Only accounts with reports from diverse, reputable users trigger reviews. (2)Transparency: Publicly share anonymized stats on resolved reports to build trust. (3)Hybrid approach: Combine reports with existing AI detection and manual checks.
These are just some criteria that popped up as I read through your post, I'm just building a vague concept on top of your solution.
To summarize, community vigilance paired with smarter automation might strike the right balance. Any thoughts, guys?
Is this an ai-generated response to an article about cheating using llms... (editing to remove the em-dashes is pretty funny tho)
Great observation, man! I've been chatting with DeepSeek quite frequently to discuss the concept of an integrated report system, so the middle numbered list section of the previous message is generated by DS and conveniently transferred here. Btw, I find it a good idea to discuss some concerning problems with AI-Tools, as long as the intent is positive.
I hope that we could be a little bit more open-minded about AI-usage in the blog pipeline... we could try to get some inspiration for discussions like this.
Of course, I completely agree that using AI for CP is completely unacceptable (at least don't use it while a contest is still running).
But it seems like that there are so many blogs about cheater-reporting systems but we're not really going anywhere. So I was wondering if discussing with AI might trigger some new ideas and build our community.
I just started high school and CP is one of my hobbies, perhaps I am not mature and credible enough to share my thoughts on AI yet — I'm here to learn, so please correct me.
We are just talking casually so you don't have to have correct grammar and correct spelling with fancy words just talk normally and give your thoughts as of we are talking face to face and this would be 100% more convenient for both of us + One point I disagree with you on is "at least not while a contest is running", if you practice with AI then you won't learn anything and would just get used to AI so you will have 2 paths either you will always relay on AI and then use it contests or you will feel like all the problems are so hard and you're not improving so you would lose interest so AI still has nothing to do here in an acceptable way
Responding with AI is craaaazy. but anyways definitely not everyone would just be able to report everyone you would need to be a trusted member(maybe with contests or from other trusted members or something) and when the cheaters get flagged they can get a different set of problems and not be considered when calculating preformence and the problems would be designed to catch cheaters(maybe problems that a human would always solve it in a way AI won't consider easily)