PDA

View Full Version : Last human to defeat a top-level computer/program in a serious controlled game



Kevin Bonham
29-03-2012, 12:12 PM
As these man-vs-machine contests that were common several years ago are no longer held, I'm just wondering who will go down in history as the last human to beat a (more or less) top-level program of the day under regulated long-play conditions and without odds will be.

The latest I can find is Ponomariov beating Fritz in 2005 in the Man vs Machine World Team Championship. Does anyone have any later examples than that?

Max Illingworth
29-03-2012, 09:24 PM
Correspondence player vs. computer, email 29.03.2012.

Kevin Bonham
29-03-2012, 09:53 PM
I'm disqualifying corro - it is virtually impossible to know the human isn't using computer assistance.

Denis_Jessop
31-03-2012, 12:25 PM
I'm disqualifying corro - it is virtually impossible to know the human isn't using computer assistance.

Or that the computer isn't using human assistance? :D

DJ

Tony Dowden
08-04-2012, 01:36 PM
I can beat my computer every time ;)

Adamski
08-04-2012, 02:32 PM
I can beat my computer every time ;)
Ok - what one do you have?

Capablanca-Fan
09-04-2012, 01:22 AM
It's certainly a worry that good GMs lose even when the computer gives pawn odds (http://www.chessbase.com/newsdetail.asp?newsid=4822), although 2700+ Milov still had a small edge at pawn and move odds as well as exchange odds (http://www.chessbase.com/newsdetail.asp?newsid=4919). But knight odds was too much even for an FM, who crushed Rybka 4-0 (http://www.digichess.gr/articles/42-rybka-232a-vs-fm-john-meyer-2284).

Igor_Goldenberg
23-05-2012, 03:34 PM
I can beat my computer every time ;)
With leg kick or right hook?

Capablanca-Fan
12-05-2017, 01:39 AM
Kasparov opines that the main advantage that computers have over the best human players is constant vigilance; humans can't help relax a little, even if it's for one move per game that's inaccurate. Against other humans, it usually doesn't cost, but a computer maintains machine-like steadiness.



https://www.youtube.com/watch?v=1pEq3ikThh8

Joschobam
18-10-2020, 05:42 AM
Ponomariov was the last win that I found,too.

But I think that not many super GMs officially played against programs since that time.

I think a Russian (Smirnov?) also won against a program at this time. But I didn't find anything more about it.

MichaelBaron
18-10-2020, 03:39 PM
People beat (occasionally) computers all the time - the question is: how strong are the programs they defeat (as the strength varies) and if they win one game out of 200...how much does it count for?