qasxcvgwe's blog

By qasxcvgwe, history, 5 weeks ago, In English

I have deeply reflected on my actions, and I realize that using AI in rated competitions under any circumstances is inherently wrong. I acknowledge that this behavior is inappropriate and regret posting a video on social media to boast about using AI. I promise never to use AI in algorithm competitions again.

After the release of OpenAI's new O1 model, I repeatedly attempted to use it for coding, particularly on Codeforces, where its effectiveness was especially noticeable. Taking Div2 as an example, after multiple attempts, I found that the O1 model could achieve a ranking of around the top 500 (at least in terms of the number of problems solved, no one in our university could reach this level). For an student who had never achieved such a milestone, this temptation was considerable. Therefore, I immediately signed up for a competition, aiming to use AI to solve problems as quickly as possible.** At the 42-minute mark, I solved problem E1 and climbed to the tenth position**. This left me both shocked and excited. Although I made no further progress on subsequent problems, my ranking remained high and increased significantly. In this state of excitement, I posted a video boasting about using AI. However, at the time, I did not understand the implications of my actions. In hindsight, it was truly foolish and ignorant, and I formally apologize to everyone.

Looking ahead, AI will undoubtedly enter algorithm competitions in large numbers; this is just the beginning. The fact that merely a few uses of AI models had such a significant impact indicates that AI's future advancements are immeasurable. In terms of code, it is completely impossible to determine whether it was written by AI because, at least on the Codeforces platform, 90% of the competitors, including myself, are not as strong as AI,AI is so intelligent that the code it writes is now almost indistinguishable from that written by humans. It can be said that in the next one or two years, online math, programming, and other competitions may lose their original seriousness. The reason AI did not have such a great influence before was that it was not powerful enough, but now, with the release of O1, the AI field has ushered in another qualitative leap.

  • Vote: I like it
  • -56
  • Vote: I do not like it

»
5 weeks ago, # |
  Vote: I like it 0 Vote: I do not like it

Auto comment: topic has been updated by qasxcvgwe (previous revision, new revision, compare).

»
5 weeks ago, # |
  Vote: I like it +8 Vote: I do not like it

"At the 42-minute mark, I solved problem E1"

Aww, it seems that deep inside you imagined that you did it. That's kinda cute. Delusional, but cute ❤

Most people are not cheaters and they sincerely hate cheaters, which is perfectly understandable. Cheaters are stealing their success, which they worked very hard, often for years, to achieve. Their wrath is a call for higher justice.. I'm pretty sure that all that negative energy does work, especially when you know that you're a cheater, and it's all directed at you and is actually deserved (all the cheaters on CF do know that). If I were a cheater, I would take extremely good care of my health from now on.

"online math, programming, and other competitions may lose their original seriousness"

It may be very easy soon to use AI to predict by rating, the total amount of solved problems across different platforms, and characteristic personal attempts to solve each problem in different years, who is a cheater and who is not. Imagine that. You can create new alts and fool some humans, but will you be able to fool AI? That's a whole new level. And everyone who cheated at least once, for fun or to boost their ego, will be placed lowest in the ranks, only because they can't be trusted 100%.

Just imagine that what cheaters are doing right now will be used against them later.. o.O And then again, all those health problems...... lol

I have a feeling CF will be fine.