Predict AlphaGo on Future of Go Summit in Wuzhen, China

[pookpooi]

But not so fast! Lee Sedol think he can beat AlphaGo if he received two stones handicap.
News in Korean language
http://news.chosun.com/site/data/html_d ... 00300.html
Top pro received handicap stones from AI is very unthinkable situation from two years ago, but as Fan Hui said, anything is possible now...
My conclusion: MCTS made five stones jump from 7k to 2k and six stones improvement due to several fine-tuning to 5d. Deep Learning made four stones jump to 9d and another five stones jump because of better Deep Learning algorithm.
End of computer go arc in go history

[daal]

Code: Select all

Jan 2016:  Who will win: Lee Sedol or AlphaGo?

Lee Sedol crushing victory (5-0) 	   26% 	 
Lee Sedol comfortable victory (4-1) 	31% 	 
Lee Sedol close victory (3-2) 	      14% 	 
Too close to call 	                   9% 	 
AlphaGo close victory (2-3) 	         4% 	 
AlphaGo comfortable victory (1-4) 	  10% 	 
AlphaGo crushing victory (0-5) 	      5% 	 




April 2017:  Who would win?

AlphaGo win against both Ke Jie and Chinese team 	        87% 	 
AlphaGo lost against Ke Jie but win against Chinese team 	4% 	 
AlphaGo win against Ke Jie but lost against Chinese team 	6% 	 
AlphaGo lost on both event 	                              4% 	

Not only is AlphaGo stronger, our predictions have gotten better too

[Tryss]

To be fair, we had very little information at the time of this poll. I must say that I voted "to close to call", as "too little info available" was not a possible answer. Now we know that AG is too damn strong