The fourth of five matches between Google s DeepMind AlphaGo and world champion Lee Se-dol has concluded and after three back-to-back defeats, Se-dol has finally defeated Google’s AI. Although AlphaGo has already won the five match tournament, Se-dol has chalked up his first victory against the AI. Also Read - Happy Father's Day 2021: Google Doodle wishes dads with cute pop-up greeting card
Lee Sedol wins game 4!!! Congratulations! He was too good for us today and pressured #AlphaGo into a mistake that it couldn t recover from Also Read - Google might be working on a 'Find My' network clone for Android usersAlso Read - PUBG New State receives over 17 million pre-registrations as closed alpha testing ends
Demis Hassabis (@demishassabis) 13 March 2016
Demis Hassabis, founder of DeepMind later tweeted that machine s loss was due to Se-dol s clever move 78, which pressurized the AI in committing a mistake that turned out to be fatal. Although AlphaGo has claimed overall victory in the tournament, Google will donate the prize money of $1 million to organizations such as UNICEF, STEM and Go societies.
The final board position can be seen below.
Google s AlphaGo combines two techniques of artificial intelligence to try master the complex game of Go. It learns with Monte Carlo Tree Search to simulate millions of games and look for outcomes and also learns from those to simplify its game strategy. AlphaGo s win against Se-dol, a professional Go player ranking second in the world is indeed impressive.
While DeepMind AI has been designed only for one task, there are other AIs that have been built specifically for games such as jeopardy and chess.
The fifth and final match in AlphaGo series is set to take place on March 15 at 9:30PM IST, which you can watch it live on YouTube (video link embedded below).