The artificial intelligence industry is booming. According to an analysis report from research institutions, before 2010, there were fewer than 20 companies involved in artificial intelligence and services in China, and this number has soared to 709 last year. From the performance of the capital market, as of November 2016, a total of 1,485 companies related to artificial intelligence technology worldwide had a total financing of 8.9 billion US dollars. Figures from market research firm iResearch show that the compound annual growth rate of China's artificial intelligence market will reach 50% in the next five years.
Artificial intelligence is striding forward toward tomorrow. Wang Feiyue, director of the State Key Laboratory of Complex Systems Management and Control, Institute of Automation, Chinese Academy of Sciences, said that artificial intelligence has transformed our era from “big law small data†to “big data small lawâ€. “The ability of knowledge automation will become the future society. The most important strength."
However, is our life really changed by artificial intelligence? How much can Chinese companies share in this "big cake" that appears almost out of thin air?
Black and white nineteen road change
Artificial intelligence has made a revolutionary breakthrough, and the concern in the field of Go is the epitome of its rapid development.
Let us turn our attention back to Go. This ancient activity has been "intimate contact" with artificial intelligence in the past year, which is a microcosm of the trend of the times.
For example, the growth rate of a thousand miles. When AlphaGo and Li Shishi, who were versioned "V18" in March 2016, played the game in the fourth set, "the mistakes that amateur players would not make", the human players also had the chance to win in the first and fifth sets. However, after 9 months, the V25 version of the AlphaGo "Master", which was born at the end of December, did not give any chess mobile phone games, and even had two chess games. Because of the mistakes made by the engineers, the "sliding marks" appeared, and after the wrong hand, the artificial intelligence Still winning. Although professional master Hu Yaoyu said that these 60 discs are fast in 20 seconds or 30 seconds, it is definitely more beneficial to artificial intelligence. But most professional experts believe that even if artificial intelligence makes the first humans, winning will still be difficult.
For example, the application market is hot. AlphaGo is not the only artificial intelligence involved in Go. Active in the game of the game on the wild fox, there are also rumors from Tencent's "extreme art" and "criminal days", they are equally powerful, the winning list even includes top professional players such as Coulee and Ke Jie. On the eve of the November 28th, 2016 Nongxin Cup competition, the artist even won an amazing five-game winning streak in the chess game with the first Korean man, Park Ting-soo. After the emergence of the Master, a new artificial intelligence master "Jilong" appeared on the wild fox. Outside of China, Japan’s artificial intelligence go “DeepZenGo†lost 1:00 in the 1:2 battle with veteran Zhao Zhixun’s “Electric King Warâ€, and Japan announced that it will host a new world competition, inviting China, South Korea and Japan. One representative player and Go artificial intelligence compete for the championship.
For example, guidance and change in traditional industries. After the "human-machine wars", in the top professional events, the opening layout of the AlphaGo is common, and human masters began to learn from the ideas of artificial intelligence. The game software ZEN has also released a new version, which is used by chess players and event interpreters to judge the game situation and even predict the next move of the players. Hu Yaoyu said: "Artificial intelligence has made a revolutionary breakthrough. The future as a tool software can help professional chess players to play an important role in this." Coule even thought, "Because of the emergence of the Master, many of us think The truth of Go has been changed."
There is energy and application, which is the reality and future that artificial intelligence faces.
From perception to cognition“Perception†makes artificial intelligence have human eyes and ears, and “cognition†makes artificial intelligence really start to make decisions through reasoning like humans.
do you know? The customer service that Taobao and Tmall serve for you is probably not a real person, but an artificial intelligence robot called "Ali Xiaomi". Last year, "Double 11" and "Ali Xiaomi" received more than 6.32 million consumers, which is equivalent to 52,000 customer service and 24 hours of continuous work, accounting for more than 95% of the total number of customer service.
"Ali Xiaomi" works by judging the problems encountered by consumers through speech recognition and semantic understanding, and then finding the answer from the knowledge base. In the view of Li Kaifu, chairman of the Innovation Workshop, a series of applications called speech recognition, semantic understanding, image recognition, face recognition, etc., called “perception†and solving “what is†are the first steps in artificial intelligence applications. .
Among the artificial intelligence "perception" technology, speech recognition is one of the most mature applications, including the video flight input method of the University of Science and Technology, the voice recognition of the notes, and the "echo" smart speaker of Amazon, which are already quite mature products. . The intelligent chat robot called "BOT" has become one of the hottest concepts. "Ali Xiaomi" is such an application. It can retrieve answers and complete instructions according to people's voice, while Microsoft Xiao Bing and Apple Siri Voice Assistant The BOTs that combined cloud computing and big data with NetEase Seven Fish have also made great progress.
Followed by image recognition and face recognition, in 2016, the Alipay login page and the large amount of money transfer on the mobile phone side of China Merchants Bank introduced face recognition verification. Baidu put the face recognition gates at the venue of the World Internet Conference. Lin Yuanqing, head of Baidu's deep learning lab, said that at the end of 2015, the image recognition rate of Baidu's brain was around 92.0%, but now it is as high as 99%. "This is even higher than 95% of fingerprint recognition." The data shows that voice and image recognition technology has become the most widely used in the domestic artificial intelligence market, accounting for 60% and 12.5% ​​respectively.
“Perception†is to make artificial intelligence have human eyes and ears, and “cognition†makes artificial intelligence really start to be like human beings and make decisions through reasoning. AlphaGo is amazing because it can handle complex problems like Go and give solutions. "Inference is the difficult point that artificial intelligence really has to face. The problem solved by AlphaGo has at least a clear victory, and more decisions are actually better, and there is no absolute right or wrong." Alibaba Cloud scientist Min Min told reporters.
From the application point of view, smart investment, intelligent search and driverless are the three hot spots of artificial intelligence on the "cognitive" level. Smart investment is a combination of artificial intelligence and investment consultants. In this field, well-known platforms such as Wealthfront have appeared abroad, and applications such as Latte Finance and Qianjing have appeared in China.
In terms of intelligent search, Sogou CEO Wang Xiaochuan told reporters that the future of search and input methods will move toward intelligent question and answer and automatic question and answer. "No longer let users summarize keyword input, but let search engines really talk like people. The same, know what the user thinks, and answer questions for them, and assist decision-making." He also revealed that Sogou has spent 9 months and spent more than 40 million yuan to develop such a smart question and answer robot.
Unmanned driving combines the functions of artificial intelligence in the two phases of “perception†and “cognitionâ€. According to data from market research institutions, the market value of self-driving vehicles may reach 42 billion US dollars by 2025. By 2030, the market share of smart cars with unmanned driving and partial manual driving functions is expected to exceed 50%.
Consolidate the foundation of developmentDespite the rapid development, artificial intelligence is constantly being applied to market products, and it needs to constantly consolidate the foundation of future development.
For ordinary users, the rise of artificial intelligence is the result of a series of "star products" market education including last year's Spring Festival Evening robot dance, AlphaGo, etc. In the industry's view, artificial intelligence is actually "big data + strong computing." + New algorithm", its rapid development is also coming from the strong support of these three aspects.
In terms of data, Yang Xu, global vice president of Intel Corporation, said that in the new era of intelligent interconnection of all things, more and more connected objects are becoming more and more intelligent, and thus a torrent of data is generated. "In the current 'network of people' era, each person generates an average of 1.5 GB of data per day, and we expect that there will be 50 billion connected smart devices by 2020, in the era of 'Internet of Things', a driverless Cars can generate 4000GB of data a day, and these huge and complex data provide the basis for artificial intelligence learning."
In terms of computing power, Tao Dacheng, chief scientist of AI Intelligence, admitted that in 2014 he studied using a camera to capture and calculate the human body posture. "At the time, it was estimated that 10 seconds of video required the best server to calculate for one hour. Now only one piece A better graphics card is fully capable of real-time calculations. The increase in computing power allows artificial intelligence to truly handle massive amounts of data."
More important is the improvement of the algorithm. Deep learning, neural networks, Monte Carlo search... they change the way artificial intelligence is thought. Wu Enda, chief scientist of Baidu, said that neural networks and deep learning are the key to breakthroughs in artificial intelligence. "Under such a model, the more data you absorb, the smarter artificial intelligence becomes."
Despite the rapid development, as a new cutting-edge technology, artificial intelligence is constantly being applied to market products, and it needs to continuously consolidate the foundation of future development. From the recent trend, on the one hand, the flow of talents from schools to the industry has become a wave. In October last year, Russ Salkhdinov, an associate professor of machine learning at Carnegie Mellon University, joined Apple and was responsible for the artificial intelligence technology development team. In November, Li Feifei, director of the Stanford Artificial Intelligence and Computer Vision Laboratory, joined Google. In China, Ubisoft also announced that the University of Sydney professor Tao Dacheng joined as the chief scientist of artificial intelligence, and in the field of robot motion control, Tsinghua University professor Zhao Mingguo carried out in-depth cooperation.
On the other hand, in terms of patent reserves, Chinese companies have also begun to come to the fore. According to statistics from market research institutions, the number of patent applications for artificial intelligence in the world, the United States, China, and Japan rank among the top three, and the three countries account for 73.85% of the total patents. The fourth largest number of German artificial intelligence patents is only 27.8 for China. %.
However, despite the rapid development of artificial intelligence, this technology is currently not perfect. Chu Min said frankly, "Artificial intelligence is just getting started. Nowadays, many applications are artificial intelligence for artificial intelligence, and the market is relatively impetuous. Artificial intelligence can't just stay in concept and must be used to solve actual needs."
3V Lithium Fluorocarbon Series Batteries
3V Lithium Fluorocarbon Series Batteries use fluorocarbon material as the battery positive electrode. The fluorocarbon material has high thermal and chemical stability. It does not decompose at high temperature ≤600℃, and does not crystallize at low temperature. The battery operating temperature range can reach -40~125℃ ; Its chemical stability ensures the safety of the battery, so that the battery has a higher safety performance when short-circuit, collision, and extrusion, and has the characteristics of explosion-proof and spontaneous combustion. Our company uses self-developed electrolyte to make the battery life more than 10 years.
Our BR series button batteries are conventional high and low temperature resistant button batteries, and the working temperature is -40℃~+85℃
Accepts customized upgraded version of high and low temperature resistant button battery, working temperature is -40℃~ +125℃
Battery application range: It can be used in fields that have strict requirements for high and low temperatures and high energy density. For example, automobile tire pressure gauge (TPMS) battery, industrial control motherboard battery, computer motherboard battery, smart instrument battery, oil field drilling platform emergency equipment power supply, marine life-saving flasher, implantable medical battery, etc.
3V Lithium Fluorocarbon Series Batteries,Cells Coin Cells For Computer Motherboard,Wide Working Temperature Rang Coin Batterise,Coin Cells Long Lasting For Toys
Shandong Huachuang Times Optoelectronics Technology Co., Ltd. , https://www.dadncell.com