Artificial intelligence talk: 6 minutes after reading this article you can be half an AI expert

Speaking of artificial intelligence, what some people see is the dawn of hope. In others eyes, there is anxiety and caution.

In 2014, Elon Musk, the founder of Tesla and SpaceX, publicly stated: “I need to stare at every move of artificial intelligence because I think it is potentially dangerous.” In the same year, Britain was famous The physicist Stephen Hawking pointed out in a machine-like voice: "The development of comprehensive artificial intelligence may lead to the loss of humanity.

However, in the vast majority of people's impressions, artificial intelligence still stays in film and television works.

What does artificial intelligence mean to us? As a cautious optimist, Cheetah’s global think tank hopes to use this article to explain things in a simple and easy way to you.

The full text of this article is about 5,000 words. It takes about 6 minutes to read. If there are any descriptions or mistakes in the text, please feel free to comment on the readers' comments. The think tank’s friends are looking forward to communicating with you.

What about Part1 artificial intelligence now? - Fruitful and far-reaching

From the 90s to the present, artificial intelligence has once again experienced a golden period of development, with numerous achievements in all aspects, especially in the area of ​​private networks, which has ushered in explosive results.

1.AlphaGo: a milestone but not a leap

Let's take the recent example. In 2016, the Go program AlphaGo developed by Google’s DeepMind beat South Korea's famous player Li Shishi and defeated humans for the first time on Go.

Humans have been defeated by AI on projects that are least likely to be defeated. This can be a milestone in the development of artificial intelligence. However, it can be said that the overflying of artificial intelligence is also somewhat lacking in energy. Artificial intelligence is still far away from us.

Compared to the dark blue that defeated human beings in chess in 1997, AlphaGo is undoubtedly an improvement in terms of the advanced nature of the algorithm and the depth of computation.

Comparing the Monte Carlo decision of chess and Go, Go requires more complex and in-depth calculations

Unlike Deep Blue's procedural calculations, AlphaGo has incorporated deep learning and self-play in its development, and it has become a bit like human beings will continue to learn and evolve. AlphaGo on the board wins over humans. AlphaGo's strategy is based on his opponent's strategy to create a strategy that is beneficial to him and completes his victory. In Li Shishi's game, the AlphaGo advantage even shows the ability to abandon partial interests and thus maximize the overall benefits. The current GoRatings ranked No. 1 Ke Jie was amazed after watching AlphaGo's game. This is not like playing a program at all, but more like a human player.

As of October 2016, AlphaGo ranks second only to Ke Jie, ranking second in the world, and AlphaGo was born just ten months later. AlphaGo is second only to Ke Jie. The "GoRatings" world Go ranking based on the ELO algorithm is a recognized and authoritative method for evaluating the current level of chess.