Google’s DeepMind algorithm edges human Go champ in man v machine chess battle
Google’s DeepMind Go playing algorithm has notched up an early win in the opening match of a five game tournament against reigning champion Lee Se-dol of South Korea.
A visibly nervous Se-dol conceded defeat three and a half hours into the bout after his silicon rival outwitted him on the board, further eroding humanity’s smug superiority over technology in the process.
The occasion is now set to go down in the history books as the moment computers mastered one of the oldest (and most difficult) board games known – 19 years after the defeat of chess champion Gary Kasparov in 1997.
Experts had predicted it would take a decade or more for programmers to build software capable of mastering Go due to the sheer volume of possible moves, which discount a brute force of assessing each move in turn.
Instead DeepMind founder Demis Hassabis devised a machine learning algorithm which allowed the code to effectively play against itself and learn from its mistakes to gain understanding of strategy and moves without the need to hard wire these in.
Following his loss Se-dol said: “I admit I am in shock, I did not think I would lose. I couldn't foresee that AlphaGo would play in such a perfect manner. I in turn would like to express my respect to the team who developed this amazing program."
Not all is lost for the human race however as the best of five rules mean it is still possible for Se-dol to pull it back over the course of the week.