From March 2016 the AlphaGo chess game in a 4-1 victory over Li Shishi has been three years from now. In recent years, Google and other large companies to contain, "artificial intelligence" common word. With the end of the man-machine war, "artificial intelligence" was also detonated outlet to the extreme. Everywhere you can see the "smart" project: intelligent voice, intelligent speaker, a smart phone robots, smart watch (bracelet), a variety of smart home and so on.
Today, artificial intelligence is the product of "deep learning" breakthrough (no wonder so difficult to squeeze algorithm section). Is essentially deep learning algorithm for neural networks, back in the 1940s proposed the neural network and reach the research boom in the 1980s.
If you want to correct deep "deep learning" algorithms are merely the hidden layer neural network adds a just, in essence, is a neural network. The "neural network" algorithms, the mathematical basis is dependent on the probability. Simply put, the speech recognition algorithm to find the closest text in the corpus; autopilot algorithm calculates the safest route; when AlphaGo chess, winning is always calculated maximum of Lazi and so on.
If only to look at the algorithm level artificial intelligence, artificial intelligence so now there is only used to upgrade the algorithm level, compared to the 1980s is no difference between the core essence of the above, do not have the ability to think independently, so why once artificial intelligence and now the difference is so big? Such as telephone robots, human-computer interaction in the customer's experience is a growing trend to replace? The key point is that big data.
Trace the origin of words, as early as 1980, the famous futurist Alvin Toffler enthusiastically the "big data" this concept was put forward in his book "The Third Wave" in, and called it "third time cadenza wave, but until June 2011, the world's leading management consulting firm McKinsey & company was issued with a large data reporting, this is the first practical application of big data.
Now, the "big data" is the word often and "artificial intelligence" mentioned together. It also has a very tacky name - "many, many of the data." For example, user data (web pages, users browse or purchase items, and even some of the content of chat software), traffic data, weather data, such as stock market data, along with increasing the number of Internet users surged storage and server computing power, storage and analysis "many many data" a reality.
Because commercial propaganda above needs "a lot of very data" the term, can not and "big data" in the name comparison.
"Big Data" is coming from the English "big data" literally, if this concept was first invented from China, it could called "macro" Is it not the same sense of style.
According to industry recognition now, AI has three levels:
First, the operation of intelligent
fast computation and memory storage capacity.
Second, the perception of intelligent
visual, auditory, tactile and other sensory ability. Humans and animals are available, able to interact with the perception of the natural world through a variety of intelligence. Autonomous vehicles, is through laser radar and other equipment perception and artificial intelligence algorithms to achieve such a perception of intelligence.
Third, cognitive intelligence
popular talk is "capable of understanding would think." Humans have language, only the concept, only the reasoning, so the concept of consciousness, ideas are all intelligent human cognitive performance.
In fact, the top two smart 1 and 2 are implemented in the software level, the human-machine interface with each other to do more easy to use, but there are still gaps from a real smart. But the commercial power is terrible, you can change the meaning of a word! Now "smart" should be almost the same as "easy to use", if replaced by a product easy to use, perfect sense, and better understanding, but not the "smart" word only tall, there is no "intelligent" two word more commercial propaganda value only.
Similarly certain furniture, under the banner name of artificial intelligence, in fact, answer the following procedure fixed frame type, even the highest technology works only with sensor lights similar principle.
Know the last few years of "artificial intelligence" is the business vocabulary, but really not much we can do. After all bigwigs play the game, as Chigua people we can not get people to forget the "artificial intelligence" is the word. However, we strengthen what they can do on the vocabulary of resistance, in the purchase of goods if the word "artificial intelligence" home like appearance, have to see if it really is their own needs, just hit this tag .
From the Turing machine - von Neumann architecture now can be said that no breakthrough nature. If you want to insist that progress, and progress should be the 5G chip improves the AI's computing power, it is like a smart phone robot mentioned above, the robot really easy to use a phone or look at the data and computing power system.
Although this "artificial intelligence" craze a little misnomer, but IT companies give some real help is especially outbound type of business, it can bring no small convenience. This is a good time, at least we have hope, and other breakthroughs in basic physics. Complete quantum string theory or M-theory to create a new branch of mathematics. Provide theoretical support, brain biology breakthrough for the realization of intelligent algorithm is different probabilities. Thoroughly understand the structure of the human brain, why would generate awareness, and whom to establish accurate mathematical model. Breaking the Turing machine - appear von Neumann architecture computer system, when artificial intelligence is the real time to achieve it!