Where is the next vent of artificial intelligence?

In 2016, one of the hottest topics in the tech circle was the human-machine competition between AlphaGo and the world's Go master. The "inconceivable" of the machine "brain" over humans has also made artificial intelligence (ArTIficial Intelligence) a hot spot for industrial investment. Will artificial intelligence become the "strongest brain" in the future to defeat humanity? Which industry will the next turmoil in the artificial intelligence market be blown to?

Recently, the China Artificial Intelligence Society held the 2016 China Artificial Intelligence Industry Conference and the 6th “Wu Wenjun Artificial Intelligence Science and Technology Award” in Shenzhen. Frontier scholars with artificial intelligence circles say that the "intelligence" of "artificial intelligence" is actually not "high", but its "reverse heaven" learning ability is still worthy of human learning and reflection.

Advantages - more and more intelligent than human beings

"In fact, the learning efficiency of the machine is not high." The fourth paradigm CEO Dai Wenyuan introduced to the reporter that the biggest advantage of the machine is that he can "learn" through massive data: "For example, let a person know a cup, I tell people 'This is a cup', people can recognize it at once, but if you have a machine, you may have to remember thousands of cups to recognize the 'cup' object. Its advantage is that it is more diligent than humans."

"It's like a machine is a child of IQ 20, but it is ten thousand times harder than you, and it will last better than you. But if there is a limit on the amount of data, so that there is no way to study hard, then it may not reach humanity. The level. "Similarly, the more times the "Alpha Dog" plays, the better the performance is based on this.

“Artificial intelligence has recently been a very hot word, from playing games through artificial intelligence to playing chess through artificial intelligence, from driverless to intelligent robots. These are all artificial intelligence. Artificial intelligence suddenly reaches a very high level. The height of the game is actually related to the data." Intel China Research Institute Dean Song Jiqiang also said that the richness and perfection of data resources has made artificial intelligence a big step forward.

"At present, we are in a flood of data. By 2020, there will be 50 billion smart devices connected to the Internet." Song Jiqiang said that everyone in the car, hospitals, shopping malls and other daily life scenes are in The massive use of smart devices that produce a wide variety of data has prompted researchers to have better data to "train" artificial intelligence algorithms.

“At the same time, these data can also allow us to generate greater value-added value through trained artificial intelligence algorithms.” Song Jiqiang said: This is why artificial intelligence has attracted so many researchers and companies to contribute resources together. How to grasp the application of artificial intelligence is very related to how to manage the flood of this data.

Development - Big Data is fuel deep learning is the engine

"Now, (industry) widely believes that artificial intelligence is the next wave of technology, and one of the important points is the revival of deep learning." Professor Lin Wei, chief research and development director of Shangtang Technology, told reporters: "If we regard artificial intelligence systems as If it is a rocket, we think that big data is fuel, and the algorithm of deep learning is the rocket engine. We also need high-performance technology to use the rocket's outer shell to explode this process." He told reporters that compared to other machines. Algorithms or tools, deep learning algorithms can give better play to the advantages of data, and the prediction accuracy is higher.

"Today is no longer the era of Amazon, Google or BAT. Today, there are actually more companies with data." Dai Wenyuan introduced a case to the reporter: "There is a joint-stock commercial bank in Shenzhen, they have a lot of marketing data, they need To accurately identify which of their customers are staging customers, we help them better identify them based on their data. In the past they also did marketing, but their model dimensions were only over two hundred, and we passed the machine through data. To help them raise the dimension to 50 million, from 200 to 50 million in fine marketing, and even in which we can help them discover some business rules."

However, it is not easy for artificial intelligence to fully bloom in various industries, and it also faces the dilemma of “difficult to land”. "Artificial intelligence is really a thing in the industry. It is not a scientific matter or even a pentathlon." Dai Wenyuan believes that artificial intelligence needs five conditions in the industry, that is, possessing data, clear business definition, and understanding algorithms. Learn how to design architecture and know how to design products.

Dai Wenyuan said: "What we have to solve is to reduce the preconditions, solve the problems of scientists, solve the problems of computing resources, solve the problem of external feedback, and let a company as long as there are IT experts, ordinary business experts, can make the model Come out and let AI reach your hand."

Next air outlet: automatic driving, sharing economy is optimistic

Many participants told reporters that if the Internet is approaching the end, the artificial intelligence has just begun. "The car of the future is the platform of artificial intelligence, and almost integrates various technologies." Huang Mingming, the founding partner of Mingshi Capital, said that the shared modes of travel such as Mobike and Didi are only in the field of intelligent transportation. Halftime", "The second half" must be driving automatically.

Zhang Jun, senior partner of Cybertron China Investment, said that the year before was hot wearables and smart homes. Last year was VR and drones. This year, Mobike and OFO became popular. He stressed: "As an investor, I would rather invest in a team that lacks technical skills, but who knows the customer and the market, and will not choose a team that is top-notch but not business."

Three Phase Ammeter

Three Phase Ammeter,Voltage Ammeter,Three-Phase Power Meter,Power Instrument

zhejiangjinyidianqiyouxiangongsi , https://www.jooeei.com