- Google subsidiary DeepMind’s AlphaGo beats professional Go player
- Chinese Game Go allows for 10,171 layouts, compared with 1050 combinations for Chess
- Software uses artificial neural networks to learn the game and has a database of 30 million moves
- Beat European champion Fan Hui in five games of a formal series
- AlphaGo mimics how humans think – deciding between top scenarios
- Founder of DeepMind Demis Hassabis hopes that the algorithm can be ‘extended to help us address some of society’s toughest and most pressing problems, from climate modeelling to complex disease analysis’
- If this technology now exists, how long will it be until this level of analysis is incorporated into the search algorithms?
This was an interesting development in AI that happened recently. I think we are closer to to the Her reality that I blogged about a while back than I realised.