Lee Sedol wants a rematch with AlphaGo

[tchan001]

http://www.independent.co.uk/life-style ... 45831.html

[Bonobo]

Awesome how the facts change as you read on into the article …

Lee Sedol wants a rematch against AlphaGo
[..] I'd like to face it again [..]
[..] he would like a rematch [..]
[..] if AlphaGo wants a rematch, I'd like to face it again [..]

(underlining added by me)

[pookpooi]

It's the KBA that contact Demis Hassabis instantly for a rematch. But then again, I purpose a much better idea for Google DeepMind. How about a triple challenge, best of 5 match for Ke Jie, Park Jungwhan and Iyama Yuta. Like day 1 Ke Jie, day 2 Jungwhan, day 3 Iyama, so the event will cover 15 days.

[Solomon]

"I figured out AlphaGo to some degree during our last meeting."

DeepMind scientists hates him! See how Lee Sedol has finally figured out the one weird trick to defeat AlphaGo and save humanity!

[Kirby]

Bonobo is correct. He didn't explicitly ask for a rematch, and the English translated articles are gradually changing the original tone.

He said this:

신중하게 생각해봐야겠지만 지난 대국에서 알파고를 어느 정도 파악했기 때문에 알파고가 재대결을 원한다면 이른 시일 내에 치르는 조건으로 다시 대결해보고 싶다

I'd have to give it some careful thought, but from my games with AlphaGo, I've come to understand it a little bit better, so if AlphaGo wanted to play another match, under the condition that we could play again soon, I'd be interested in playing again.

So yes, Lee Sedol would probably accept a match against AlphaGo again, but under the condition that they could play soon (my guess is that he understands that AlphaGo will continue to get stronger, and it may be a different opponent in a little while).

And it's not a request that he's making to Google, but that he might accept the request, if AlphaGo (Google/DeepMind) were to request it.

Anyway, it does seem like he enjoyed playing AlphaGo, even though he lost the contest.

[Bonobo]

[..]

He said this:

신중하게 생각해봐야겠지만 지난 대국에서 알파고를 어느 정도 파악했기 때문에 알파고가 재대결을 원한다면 이른 시일 내에 치르는 조건으로 다시 대결해보고 싶다

I'd have to give it some careful thought, but from my games with AlphaGo, I've come to understand it a little bit better, so if AlphaGo wanted to play another match, under the condition that we could play again soon, I'd be interested in playing again.

Thanks for giving us a translation that’s not such a Telephone game as in that article.

So yes, Lee Sedol would probably accept a match against AlphaGo again, but under the condition that they could play soon (my guess is that he understands that AlphaGo will continue to get stronger, and it may be a different opponent in a little while).

I wonder how many millions of games played by Ali would be acceptable for him, and one than that more would be one too many

And it's not a request that he's making to Google, but that he might accept the request, if AlphaGo (Google/DeepMind) were to request it.

Well, even though I wrote that other comment above, I admit that there is some <plink plink> effect there—and how much of it by intention? It often is easier to say that you’d agree to something than to admit that you actually want it (badly). (Like we so often ask somebody “Would you like to …?” when it’s actually ourselves who want to … )

Anyway, it does seem like he enjoyed playing AlphaGo, even though he lost the contest.

Yes, this is something I can really admire, even more than winning so many games and beating so many other strong players …