Host: The match is over. We have 5 questions here for Qi Guan, our world champion.

Host: First question, how do you evaluate your performance in the match?

Qi Guan: I think I basically performed as expected, though I made some mistakes. I was not in my best form, but perhaps the best needs some good luck. It was my normal performance.

Host: You caught a cold recently. Moreover, it's match with a machine. So I'm wondering whether your performance was influenced by that a little bit?

Qi Guan: I guess I wasn't influnced by that. It's hard to say when I could be in my best form. But as for my playing style when playing with a machine, it was different from playing with human players.

Host: What advantages and disadvantages of Yixin can you see from the 2nd and 4th game?

Qi Guan: It can be observed that, Yixin has an advantage when the position is relatively simple. That is to say, when the number of stones is relatively small and the winning path is not very complex, Yixin's analysis is pretty accurate. In the 4th game, there's a relatively simple winning path. Yixin found that path in a short a time, which is very strong, because human players may not be able to, or need a lot of time to find it. I made a mistake in the 4th game's opening, which led to a quick loss.

For the 2nd game, in general, Yixin's performance was good, though one of its move had some problem. When there are lots of stones on the board and the position is relatively complex, Yixin's analysis may miss some winning paths. I don't know exactly how Yixin makes analysis, but for example, in the last game, several moves were made that neither created materials nor made connections locally. Probably Yixin thought there was no threat so it ignored my moves. However, after Yixin ignoring several moves of mine, I finally got a good position. I think in terms of that Yixin has a bit of a problem.

Host: If a human player controled Yixin to play with you in the match, how many scores do you think you would get?

Qi Guan: If the match were a human+AI vs me, the game would be very very tough for me, but it might still depend on what variations to play. For variations that have very long well studied opening sequences, I think it might still end in a draw. But for relatively complex variations, it should be tough to draw.

Host: Do you think a renju AI will beat human players one day in the future?

Qi Guan: In theory, eventually it will. However, it may take a long time or a great effort to achieve that. Moreover, the answer still depends on variations. Maybe that's the feature of renju. For variations that have very long well studied opening sequences, the game might still end in a draw no matter how strong the AI would be. However, for normal variations, it would be hard for human players to compete with if in the future an extremely strong AI could be made with human like thinking added. Of course, I think it's very difficult to add human like thinking to an AI.

Host: Many renju enthusiasts around the world followed, supported, or watched the online live. I'm wondering is there anything you'd like to say to them?


Qi Guan: Play more if you really love renju. For regions where renju is not popular, I wish you can introduce it to people around you and let them enjoy renju.

Host: Great! Thank you, Qi Guan for your speech, and thanks, everyone, for your attention to the match! See you next time!

Qi Guan: Thanks Kai Sun for inviting me to play with Yixin. It's my honor. And thanks, everyone, for your attention. See you next time!

Host: Thanks!