RSS Feed

2/02/2016

Topic Reading-Vol.1392-2/2/2016

Dear MEL Topic Readers,
Google achieves AI 'breakthrough' by beating Go champion
Google won. This time is not in a search engine or smartphone battles but a board game, Go. IT is a Chinese board game that became popular over two thousand years ago. Though the rules are much simpler than other games, there are 10 times more possible choices for each move than chess.
Google’s DeepMind division challenged the European Go champion and the program won five to zero. This was achieved by a technical idea of reinforcement learning - getting computers to learn to improve their behavior to achieve goals. In fact, it learns from its mistakes. The program now plays different versions of itself millions times, and each time it learns and gets incrementally better.
It should no longer be called computer but artificial intelligence program that now could beat human intelligence and experience and possibly assist humans to make decisions.
Enjoy reading and learning what the latest artificial intelligence manes and could mean to you.

No comments:

Post a Comment