Why does the development of artificial intelligence need the country?

If you're a movie fan, you must have seen this year's Oscar-hit movie "Imitation Games," a biography film based on the famous British scientist Alan, shows how Turing designed a process that mimics the human thinking machine to decipher the Nazi German code.

If you are a science fan, you must see Hawking, Musk's fear of a rising upsurge, musk even asserts: "We need to be extremely vigilant about AI, which is more dangerous than nuclear weapons!"

If you are an Internet addict, you will be "brainwashed" by internet companies in the past year, including but not limited to machine learning, computer vision, etc.

If you are a news fan, must recently pay attention to the two sessions, to CPPCC members, Baidu founder Robin Li put forward the "Chinese brain" plan, a time to cause media hot debate.

All of this is attributed to one word: artificial intelligence. In the past year, artificial intelligence has become the hot word of the whole Internet without any hesitation, its influence is not far from the term. But as with the industry's heated data and cloud computing, artificial intelligence is not a new concept. As early as the 50 's, Professor John Mackenzie of MIT put forward the concept of "artificial intelligence", which refers to the science and engineering of making intelligent machines. Subsequent artificial intelligence developed ups and downs. After a great heat in the 80 's, why did this concept rise again in the 21st century when humans entered the internet age?

The three technology trends make AI no longer distant

The Internet revolution brought not only the world to connect closely, the formation of McLuhan predicted the "global village", but also made previously unimaginable technology to become a reality. If you want the program to understand a spoken word as intelligently as a person, you must be able to hear all the phonemes (syllables) from each other, and when the program identifies the key technology for a picture, you need to see all the pixels in the surrounding pixel environment. In recent years, the GPU technology gradually mature, GPU is a visual game to meet the high density of visual and parallel requirements of the chip, now is Baidu's chief scientist Wunda in 2009 to use the GPU to establish a parallel neural network, in order to achieve such as recognition of the image of the target.

The vast numbers of data in the Internet age provide a large sample of artificial intelligence research, which helps to make machines more fully recognizable and learned. In machine learning, the next generation of algorithms such as depth learning to stand out from the previous machine learning model, depth learning by combining low-level features to form more abstract high-level features.

For example, to identify a face in a picture, for example, the machine begins with the edge of a face, from small to large, and then combines the elements, and finally draws the conclusion that the process is similar to the human brain's recognition of human faces. At present, from the domestic situation, Baidu in the depth of the field of learning has a large competitive advantage, especially after Wunda joined, Baidu in-depth study has gradually extended to the entire Baidu product system, it is estimated that because of this, Li has emboldened in the two sessions of the "Chinese brain" proposal.

It is under the interaction of these three technology trends that artificial intelligence technology has been the focus of the industry in the past few years. The data compiled by the CB Insight researcher showed that only two AI companies were set up in 2010 years, while 2014 increased significantly to 16. More importantly, the European and American Governments, as well as the internet giants are in the intense online AI project, such as the European Union's human Brain program, is to establish a new, revolutionary information and communication technology based on neuroscience; Google, IBM has acquired and developed the next generation of intelligent robots and smart chip; The U.S. Department of Defense's Advanced Technology Institute, known as DARPA, is pushing for a new program of human-computer interaction CWC to better communicate between humans and machines. Why are these countries or enterprises so fond of AI?

How much is the imagination of artificial intelligence?

Let's take a look at the current research field of AI:

Human-Computer Interaction (Natural language recognition)

Computer Vision

Speech recognition

These three basic research will greatly change the development mode of the existing industry, taking manufacturing as an example, with the maturity of natural language recognition and speech recognition technology, the interaction between human and machine becomes simpler, on the basis of which, industrial robots can better understand the general Workers ' password. Computer vision Technology can help the robot to accomplish the task precisely in the process of the equipment manufacturing.

The potential of artificial intelligence in manufacturing also coincides with the current industry discussion of "Industry 4.0". "Industry 4.0" emphasizes the importance of intelligent factories and intelligent robots, which need a more intelligent system to support, and artificial intelligence will undoubtedly play an important role.

If the scenario is still in the factory, the driverless car lets the artificial intelligence technology go into the ordinary family. To make cars unmanned, the first need to let the car understand the "road", based on the identification of various objects on the road-around the car, road pedestrians, roadblocks and so on, this must rely on computer vision technology for the car to deploy good "eyesight." Second, speech recognition and human-computer interaction can help the car better understand the driver's ideas, do safe driving and civilized driving.

Many people who fear artificial intelligence have repeatedly reiterated the view that artificial intelligence is to let human unemployment, from physical and intellectual victory over mankind. This is true, of course, in manufacturing, for example, as the popularity of smart machines, factories need fewer workers, but the efficiency of the factory has increased dramatically. And a large number of factory production patterns change the entire manufacturing sector transformation and upgrading, which is Germany in the name of the country to start the "Industry 4.0" and the EU's "brain plan" to promote the original intention. By the same token, the US government has also injected a large amount of money into companies or projects around AI research through DEPRA and NASA, why does artificial intelligence research need support at the national level?

Without the impetus of national power, artificial intelligence projects or fear of success

Human progress has been inseparable from the power of technological innovation, but with the continuous development of technology, can really promote the development of technology is not just individuals, companies, and become a national mission and responsibility. This is because as technology grows more complex, the power of individuals and companies cannot be fully assumed. Historically, the rise of the first industrial revolution was almost as Watt's strength improved the steam engine; the second Industrial revolution was driven by several power-related companies, but in the third technological revolution, the legacy of the "Manhattan Project" of the United States during World War II, from Bell Labs to various military companies, Become the core force that led the United States to become a global technology leader after the war.

In a Ted speech in 2013, Professor Mariana Mazzucato, from the Center for Science and Technology Policy Studies at the University of Sussex in the UK, put forward a surprising view: in the economic market, the most innovative trend in the role of the country!

Mariana Mazzucato's argument stems from a study of her rival's iphone, which she found:

"These ingenious and innovative parts of the iphone are funded by the Government, which means the mobile Internet is funded by DARPA, the Pentagon. The GPS is funded by the military Navstar (navigation Star); even Siri, which is actually funded by DARPA, has a touch screen funded by two public university researchers at the University of Delaware, with two national departments-cia (CIA) and NSF (National Science Foundation). ”

The research on artificial intelligence needs more state participation, not only because of the astonishing cost of artificial intelligence research, such as the European Union's "Brain Project", which the European Commission will support for its research in the next 10 years, The IBM Smart chip TrueNorth mentioned above is supported by the U.S. Department of Defense as much as $50 million trillion. Such research funds have already exceeded the scope of ordinary people and even enterprises, only from the national strategic level of financial support to ensure the sustainable development of research.

Secondly, artificial intelligence is a basic technology to seize the future competitive heights. China is a manufacturing country, but not a manufacturing power, in recent years, China's manufacturing industry faces the double pressure of rising labor and raw material cost, the cost advantage of traditional level disappears gradually, the transformation of manufacturing industry is imminent. As the previous article, artificial intelligence will fundamentally change the traditional manufacturing mode of production, so that China's manufacturing industry rapid transformation, at this level, China needs and traditional manufacturing power, such as Japan and Germany to compete, its importance is self-evident.

Third, the initiation of a national level of artificial intelligence project is also a new approach to the government's innovation model. China has experienced a long period of planned economy period, the government has long "occupied" innovation what how to innovate the say. However, with the change of government responsibilities, the government has begun to put some innovative projects in universities and scientific research institutions, but because of the lack of necessary market sensitivity of universities and scientific research institutions, their innovative products are often unable to be truly marketed. And China's AI program, is the state, universities/scientific research institutions, enterprises triple exerting force: the state is responsible for policy approach and financial support, colleges and universities to promote basic research and talent training, enterprises combined with market and national demand, the basic research products, and the use of different industries and industries demand to achieve the optimization and adjustment of industrial chain.

Standing at this time of 2015, human beings are full of longing and fear for the future of artificial intelligence, as Europeans did for Gutenberg's printing press in the 1400. However, we have reason to believe that, just as Gutenberg's printing ended the savage medieval civilization of Europe, and thus sounded the Horn of the European Renaissance, we have reason to believe that artificial intelligence can be the wings of human Take-off, and China, in this game cannot afford to lose, also is ready to be sent!