Блог пользователя abbcdbbc

Автор abbcdbbc, история, 3 месяца назад, По-английски

In response to the previous discussion regarding AI, I would like to offer my thoughts.

we all are doomed

Edit: I just found this blog which was hidden due to downvotes https://mirror.codeforces.com/blog/entry/133962. AI got master perf last round, we are cooked.

  • Проголосовать: нравится
  • +19
  • Проголосовать: не нравится

»
3 месяца назад, # |
  Проголосовать: нравится +5 Проголосовать: не нравится

i think the correct expression here is : "WE ARE ALL FUCKED"

»
3 месяца назад, # |
  Проголосовать: нравится +8 Проголосовать: не нравится

What we've all been waiting for has finally happened: we now have a record of an AI model participating in a rated Codeforces round (and getting a 2.2k performance on top of that)

You could argue that the problems were a little bit easier/standard than usual, and in the end problem C solution got hacked, but I don't know if that counts as cope or not

»
3 месяца назад, # |
  Проголосовать: нравится -15 Проголосовать: не нравится

What do we humans have left? Philosophy? Will AI eventually become more thought-provoking than people like Plato, Trump, and Locke?

  • »
    »
    3 месяца назад, # ^ |
      Проголосовать: нравится 0 Проголосовать: не нравится

    Let's hope AI can answer questions philosophers have struggled with, like "what is the meaning of life?" or things like "solipsism: I can be sure of the existence of only my mind".

»
3 месяца назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

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

»
3 месяца назад, # |
  Проголосовать: нравится +21 Проголосовать: не нравится

I refused to believe that chatgpt could successfully solve E1.

So I tried it myself. I independently verified it: It successfully did it, in first try (so, 60 seconds),281286880, it does not even take 8 minutes as suggested in the downvoted blog.

Then I realised I read the constraints incorrectly: that an O(n*m*l) passes (I just went for E2 in contest)

So the issue really is that this simple dp is being rated 2100 (it should really just be rated 1500), and that many humans failed to take advantage of it. Still, even in this case, the rankings/ ratings implications are massive.

  • »
    »
    3 месяца назад, # ^ |
      Проголосовать: нравится 0 Проголосовать: не нравится

    agree with this it was way too easy for E1 i would argue C was a bit harder in terms of implementing.