Hola CP-igos!
As the name suggests yet another blog an AI, cheaters etc.
Introduction
Flashback to 1.5 years back... I just joined Codeforces, all contests were fine, no cheater blogs(and no cheaters)... I used to spend some time reading interesting blogs in catalog and the "Recent Actions" where I used to find blogs on other "non-educational" blogs, be it somebody's productivity plan, somebody reached GM, some new solutions on rounds, the "if u comment under this blog" blogs etc. All this was fun.. I could spend an hour or so, reading all this and the comments underneath them
The Issue
Fastforward to now.... A round gets over.. And soon by the next day, all we have is tons of blogs on cheaters doing this, cheaters doing that, I found this, I found that and what not. And then people finding underneath blogs, seperate blogs being made for seperate cheaters, as though the cheaters have become kings, every set of cheaters with the same solution gets a seperate blog dedicated to them. Gone all the useful blogs, and all for sometime, drowned... Plus mood getting spoilt
Personally I do not know anybody who likes this, this becomes downright irritating.
Suggestion to Sort it Out
Suggestion 1
For every round we have testers who test the rounds.. we can also have testers who are proficient with AI to make the problems AI-safe. Manually checking for variable names won't last long.. Cheaters are also learning, soon the techniques used to detect cheating like "tons of comments in code", "similar variable names" etc. wont be a problem for the cheaters, they will start removing the comments, start giving weird variable names and so on. So eventually if we need to stop this for once and for all, we should have AI testers.
Suggestion 2
To all those who are still interested in detecting post-round cheaters can still post comments, but we have a separate blog made (preferably by the maker of the round), named "Round XYZ : Cheating Analysis" or "Cheating Reports"(or a better name, acutally a better name would be better I dont really like this name) as made by chromate00 like Anti LLM Evaluation Report, where the problem setters can actually inform the participants about the cheatings caught etc. And we can have comments under the blog regarding the cheating suspects they have(preferably in bold, because it might get drowned)
Well this is all I have to say.. If anybody thinks I might have made a mistake or you dont think my opinion is correct, anywhere please let me know. And if you have any more suggestions too!








I'm not active on CodeForces at all and I'm not familiar with blogs in the present or past, but irrelevant to this — considering only Suggestion 1 — cheaters are "learning" as you say, and there are probably already "smart" cheaters who will just copy the idea off of AI. This is unstoppable — below a certain difficulty margin, AI will give a solution, and cheaters will cheat — unavoidable for something subjective like code.
Nonetheless, I agree with the more general point about how there's no real use of ranting randomly and the voice and effects are probably amplified if done in one place.
We got ludo. posting a blog before GTA VI
I strongly agree with suggestion 2. However, I don't think making problems "AI-safe" in suggestion 1 is a good idea. I completely agree with this blog.
Auto comment: topic has been updated by ludo. (previous revision, new revision, compare).
I think that since you are a member of evenvalue fan club, there can be no disagreement over the fact that your opinion has to be correct. So, since you have 2 suggestions by even number theorem either both are correct or both are wrong. And since suggestion 2 is really nice, suggestion 1 must be too :)
evenvalue orz ludo orz
I have some interesting thoughts on AI cheaters.
If we have AI that are strong enough to solve, say, 2500-rated problems, why not use AI to predict a contestant’s performance in a contest (for example, their expected score)? Then the AI would make the rating changes, making individual rankings irrelevant.
I think the key issue with AI in this context is that when we see cheaters ranked above us, we get mad! (This really happened to me in my last contest when I saw 181 accepted submissions on Problem D at 1 hour and 10 minutes, and then 800+ accepted submissions 20 minutes later.)
really strange to blame more people solving the problem at 1h 30m when you yourself are one of the people who did it afterwards then?
Imho just make separate section for reporting cheaters, and prohibit it from being posted into normal blogs ¯_(ツ)_/¯
loopyposting orz