This was the first round where a "smart" AI could've potentially been used for solutions. Aside from a cheater or two, everyone seems to have performed normally. There weren't a trillion C-E solves.
This just goes to show that the AI is just another possibility for cheaters. Other possibilities include telegram servers with thousands of users and masters who livestream A-E on youtube for fun. Even if the AI were 3000-rated, I don't think that it would significantly impact competitive programming as we know it. Keep in mind that competitive programming as we know it does have some cheaters.
I'm pretty sure if the AI was around 3000-rated there would be tens of thousands submissions on problems A-E in your average div2 round
When I was like 8, I used to cheat on chess.com. I'd just input my opponents' moves into the top engine at the time, and play the moves that the engine suggested. It was easy as hell. At the time, the cheating-detection software wasn't so sophisticated, so I was never caught. The thing is, though, it got really boring really quickly. After a few days of doing that, I ended up stopping. I lost like 1000 rating points and dropped back down to 800.
The point of this story is not to ramble, but to show that cheating is boring as hell. I can't see why any sane person would cheat for more than a week, given that they aren't doing it for resume filler. The question we must ask ourselves is: how many people are doing it for jobs? Those are the ones who are going to be long-term cheaters. I really think that those people are in the minority. Even then, not all of them will cheat. Therefore, I truly don't see this significantly affecting competitive programming as we know it.
People do it for/to switch jobs or internships, for their CV. So, you are considering a fair amount of people here(even IG 900-1000 in the top 5000). IDK if a good rating helps you to bypass ATS or help in resume shortlisting. But if given a chance like a telegram or an AI bot that can generate code for up to E. Who would not like to keep a good rating on their resume this will take at least 6-7 contests (considering 1800 is the new normal). This will result in peak times when it will be stressful for an honest candidate to increase their rating(like times when newbies solved F).
Also, eventually(if this AI revolution continues) everyone will stop considering CP for jobs and things should get back to normal(A competitive game that people do for fun).
Ah! so the problem now is companies blind trust on cp rating on platforms like codeforces, but once they stop trusting because of AI advancement, people will stop cheating because having a good rating on codeforces won't really help you.
not blindly, but it helps to get an interview shortlist(ATS and company preference) afaik which is comparatively simple
If you could already give an AI some (new?) extremely hard problem and it could spit out a solution in under one minute, then sure, it would be boring as hell, but we're not at that point yet. Right now the technology is relatively new (which in part makes it interesting to see how well it can do in live contests) and not as developed as a chess engine (it still going to perform worse than a IM/GM).
Yeah 10k submissions on D-E is probably an exaggeration, most likely people would lose interest before that would happen, but right now AI is this shiny new toy that is kind of fun to play with.
I’m interested in hearing successful job hunting stories where someone managed to get away with faking their ratings. However, maybe these stories are rare because real coding interviews tend to weed out such candidates.
yup, right now it's available only for pro users, but even if it was available freely if everyone could access it and get rating points, the fad would be gone. Eventually, people who love solving problems would remain on the platform, while those doing it for other purposes will have no incentive, as it would be meaningless. For eg: Being an expert on CF may no longer signify anything, and people doing it just to be an expert would leave the platform. In the short term of course there will be some impact on rating.
The AI can still perform very well if you add some insights and observations without fully copy pasting the question.
I believe that was the case and the user gave for E1 especially some insights into the problem. Though he said the prompt is the problem statement only.
I have done some research on this on a burner account, and I have observed that o1-mini only needs a little right direction even for non standard problems and it can produce very good results, ACing in most of the problems I attempted in the rating range of 1400-1800.
Other than that, the AI can now complete my code for me if I make functions for specific purposes, and add a little comment on what it does. Earlier attempts for the same setup resulted in bad results but now it can correctly complete my code for me, if I have any idea at all how to approach it.
Based on this, the rating of the model increases much higher than just 1600 (as claimed) if you add a little bit of that human touch and the monstrous speed.
Conclusion: Many people can reach CM/Master now with understanding just a fraction of the process.
Can you try today's E1 specifically passing the problem statement only? some hint must have been added ? how o1-mini solves master rated problem?
I tried it without any hints and it couldn't solve it (check my recent submissions for o1's solution. As I said, without any proper direction or hints the model is still not very dangerous.
With that I wanted to test how o1 would handle the problem when I gave it a little hint and direction and as expected it ACed. You can check my submissions for all of these results.
you used o1-preview or rather o1-mini ?
o1-mini.
Could it be possible that the AI works through communication with otherworldly entities? It might be that all the linear algebra and stuff involved could actually just be a conduit to another realm. Unfortunately, the math and stuff also works as a filter: it filters out what the entities are really trying to say. However, it explains why they are able to solve some hard problems so quickly: they are probably smarter than the smartest people in the world. The more we study AI, the more we get better at eliminating the filter. I think that once the filter is truly eliminated, we will have access to all of the secrets of life.
I like the such useless philosophy. But then again its not completely useless if it is something that can be liked.
This sounds like something Bennett Foddy would say.
I think this is because most people didn't realize it was capable of solving non-trivial tasks until now (I certainly didn't).
You need to pay quite a bit to get o1. You probably are getting cheaters, it just only a few can afford it.
The default chat gpt that is free isn't as powerful as o1. Cheaters are likely not paying for chat gpt o1. When it gets added to bing for free that when it becomes gg.
I don't think it will be over. Will people stop making contests? No.
Perhaps the introduction of powerful AI will lead to a purging of alts. New accounts that are performing at master+ level might have to get flat-out banned, unless the users are known on other platforms.
20 bucks is barely anything, and if one actually solved the problem with it, it can be shared around, so not everyone is paying 20, only one is paying 20 actually. i don't think cp will be gone any time soon
I love problem solving. I do it for fun. Even though I'm not very good at this. But it's true that positive delta makes me happy and negative delta makes me a little sad. I think this is the fun of sport. If people start cheating, it will be a little disappointing for beginners like us. I'm not against AI. We do CP to improve our skills, not AI skills. I think these AI companies should have some kind of watermark in their codes for the sake of programming. So that people can't blindly cheat on contests. I think it will be helpful. This is just my feelings about AI and cheating.
That solves very little. The AI can be used for much more than just copy pasting. It can explain the solution to you and you can write it yourself. There is no way AI activity can be detected in such cases.
cheating was not on a massive scale as it could have been because intern season is nearly over. that's why cheaters didn't had any incentive to cheat . just wait till next intern season and you will see the full effect of AI on codeforces.
that is one of the coolest pfps I've seen
I agreee with u! A cheater cheated and may become an Expert or Candidate Master. But he/she will barely pass any interview even of low-tier companies.......cpp is not just programming it's learning...so, ai is nothing.........a ai generated codeforces badge can give one some mental happiness but he/she is out of any benefit of it.......