Explore the key behind the artificial intelligence AlphaGo to win

Artificial intelligence is constantly evolving

Since the advent of the first computer, people have dreamed of inventing and creating a computer system that can perfectly simulate or even surpass the human brain. In the past 20 years, four human-machine wars have left a particularly deep impression on people and have become an excellent driving force for the development of artificial intelligence. Of course, if you add this AlphaGo, it should be 5 times.

Why artificial intelligence AlphaGo can win, cloud computing is the key behind the scenes

1. In 1997, IBM's "dark blue" defeated Kasparov, the greatest chess player in human history.

2. In 2006, the man-machine war, the wave supercomputer "Tissot" and the Chinese chess grandmaster Xu Yinchuan battle.

3. In 2011, the “Deep Blue” fellow “Watson” successfully challenged two human champions in the American veteran quiz show “Dangerous Edge”.

4. In October 2015, the “Alpha Go” artificial intelligence program defeated the European Go champion Fan Wei by 5:0.

The development of artificial intelligence is staggering, and deep learning is indispensable. To defeat the top players of human beings, you must learn to make computers think like human brains. In this competition, Google designed two neural networks for AlphaGo, using "two brains" to solve the problem: "decision network" and "value network." In general, one brain is used to decide how to drop the current one, and the other brain to predict the final victory of the game. At the same time, AlphaGo also researched its own strategy, running thousands of Go games between its neural networks, using trial and error to adjust the connection points, and doing a lot of research work through extensive use of the Google Cloud Platform. It is this kind of super learning ability that is the key to AlphaGo's success. Through the performance of AlphaGo, I believe that everyone can also see the importance of deep learning. By establishing and simulating the neural network of human brain for analysis and learning, let The machine can think like a human being, and it can be said that it is the most promising technology for artificial intelligence.

Artificial intelligence will become more advanced cloud computing

The key technology of artificial intelligence is deep learning, and deep learning can be said to be substantial progress in the context of the maturity of cloud computing and big data. Cloud computing provides a platform for deep learning, while big data provides deep learning for deep learning. cornerstone.

Google’s senior researcher, Jeff Brain, head of Google’s brain team, said at the press conference before AlphaGo and Li Shishi that Google’s machine intelligence has brought about tremendous changes and more and more opportunities, and artificial intelligence has become A higher level of cloud computing services. Through the Google Cloud Vision API, developers can apply Google's image recognition technology to their products, and the open source artificial intelligence tool TensorFlow has become the first machine learning class repository on GitHub. Dean said that Google provides labor. Intelligence already exists in Google's cloud computing platform. Artificial intelligence for deep learning through neural networks is based on cloud computing, open artificial intelligence interfaces and artificial intelligence tools, which will become higher-level cloud computing services. It will provide more advanced features based on data analysis.

It is undeniable that cloud computing and big data play an important role in the evolutionary history of machine biology. In the future of more and more open source artificial intelligence projects and services, high-performance computing and cloud computing will also be blended.

Written at the end

Back to the game of AlphaGo and Li Shishi, artificial intelligence is powerful, but Li Shishi’s defeat can not explain what the problem is. “Humans have been threatened by robots.” “The machine will eventually replace humans.” The title of Xuan, in Xiaobian, is to pull, a game only, can only explain the development of artificial intelligence, Li Shishi bears too much public opinion, on the skill, AlphaGo's chess skills belong to the professional 1-3 segment And Li Shishi is a professional nine-segment. But the machine is a machine after all, it has no emotional factors that humans have. It won't be afraid, it won't care about winning or losing, it doesn't have any sense of mission, it won't be disturbed by the outside world. When playing chess, AlphaGo can be said to be 100% invested, and Li Shishi is too emotional, fidgeting, and uncertain. It’s not surprising that I lost to AlphaGo. "Winning and losing are human victories! Because it is human efforts that make artificial intelligence have a breakthrough now." I agree with Google's chairman Schmidt.

Brushless Dc Motor Controllers

Dc Servo Controller,Dc Servo Motor Controller,Dc Electric Motor Controller,Brushless Dc Motor Controllers

Jinan Keya Electron Science And Technology Co., Ltd. , https://www.keyaservo.com

Posted on