VR has been played bad but I hope AI is not so bad luck

Source: 歪思妙想/neihangaoxiao At the beginning of the concept of artificial intelligence, it carried the infinite imagination of human beings for the future world. This humanoid agent modeled by wisdom and thinking is a shadow of human beings from a shallow level, just like the shadow of a shadow play. Take control. But at a deep level, all aspects of artificial intelligence will inevitably become closer to human beings. Like the Tomoko in the three bodies, it is a perfect existence that enables humans to forget about hatred and aggression. At that time, it was even in the whole society. The definition in the natural world will inevitably change with the unpredictability of human emotions. The so-called human-computer relationship may be fused. However, it is this kind of feeling that is similar to the creation of a new living body, which makes our human ambition and sense of honor infinitely inflated. In particular, in the atmosphere of capital blitz and concept marketing in recent years, an illusion that artificial intelligence is within reach is created. This is almost exactly the same as last year's super hot VR technology. But the fact is that after looking at what was known as VR in the first year of 2016, it was only after a countless chicken blood fever that it was a brutal bubble burst. With Alpha Dog's frequent shots, and Baidu artificial intelligence, various news, the current state of the industry is gaining the attention of the people. But what needs to be warned is that perhaps the current artificial intelligence is like a mirage in the desert. At the same time, a large number of followers will fall to death. As the artificial intelligence scientist Li Feifei fears, excessive bubbles will only bring about unfulfilled promises and goals. Not surprisingly, the winter of capital will continue in the coming year, but the heat of artificial intelligence may also continue to rise. Does this reality and ideal conflict make many startup companies face greater difficulties? Will the artificial intelligence bubble shift from looming to large scale? Entrepreneurs and the public have excessive expectations of artificial intelligence Artificial intelligence enters the general public's field of vision, thanks to the two world-viewing match-ups. At the beginning of last year, the Alpha Terrier had abused Li Shishi. Maintaining the final dignity of mankind in the world of chess seems to have become a burning eyebrow. Less than a year later, the Alpha Dog has changed its name to “Master” and swept the country to win 60 consecutive games. An uproar. Unknown, the onlookers are admiring and have more inexplicable expectations for artificial intelligence. In fact, in the final analysis, this is not a disruptive technological breakthrough, because in terms of algorithm, human beings have long been difficult to contend with computers, and Go is precisely based on fixed rules that can exhaust the algorithm, so in essence, this may not be intellectual and The layout of the competition, there is no such thing as the dignity of the chess world. It is undeniable that Google's two advertising campaigns have done extremely well. They have not only perfectly displayed their superior artificial intelligence programs, but also affected the world's public opinion on artificial intelligence. Afterwards, with the help of Internet giants, the industry seems to be thriving. According to Wuzhen Index: Global Artificial Intelligence Development Report 2016, an artificial intelligence enterprise was born every 10.9 hours in the world, and there were 709 artificial intelligence enterprises in China. Although it is much worse than the United States' 2905, this figure is enough shocking. Recalling last year’s booming Internet broadcast platform, the peak period reached 200 or so, and artificial intelligence companies with a much higher technical threshold can actually achieve such a scale in a short time, which means that there are as many people in the field. Is it really possible to carry the corresponding commercial value behind this number? According to the relevant data on total financing, China’s investment in artificial intelligence in the past year was approximately US$1 billion (approximately RMB6.8 billion). However, companies with statistics from the top 100 in the financing rankings found that the top ten AI companies have achieved a total financing of more than 200 million yuan, while companies with more than 100 million yuan have as many as 18 companies, such as the company's C only. The round of financing reached 200 million yuan, and Yun Zhisheng's B round of financing was 50 million US dollars. He went out to ask for a cumulative financing of 75 million U.S. dollars. It is not difficult to see from these figures that most of the financing amount is concentrated on the head, and the tail-end enterprises with large bases actually receive little financial support. From this perspective, all artificial intelligence companies must have followers who follow most of them blindly. They are just like the bubbles created by this boom. They seem to be spoiled but they can only drift away from the wind and drift away. In fact, whether or not this AI feast is unworthy of its name, the market opinion created by capital and marketing alone is not necessarily a good thing. Entrepreneurs or the public, if guided by propaganda and public opinion, are extremely vulnerable to unrealistic expectations. If such over-expectations cannot be translated into reality as soon as possible, they will have serious adverse effects on the market, and may also cause many small companies to move toward it. Perish. Entrepreneurial threshold lowers, but the threshold for success is improving Luo Ji thought that the original algorithms for autopilot, pattern recognition, and speech recognition were different algorithms. However, due to the existence of deep learning algorithms, the bottom layer was opened, resulting in a lower threshold for entry into artificial intelligence. Businesses began to move into the field of artificial intelligence. However, there are also other factors that have inspired entrepreneurs to enter the confidence that the more mature artificial intelligence algorithms are now relying on massive information in the era of big data and benefit from China’s huge population base. Information acquisition does not seem too difficult, so this has even become the overall advantage of domestic startups. However, after all, the data is not all of artificial intelligence, and in the final analysis, it must rely on important breakthroughs in algorithm, otherwise it will only stay in the initial stage of artificial intelligence. And even if it is a startup that is trying to rely on big data to accumulate resources and technologies, it will also become a cannon fodder for large-scale companies in the external environment of the giants of 2017 and the cold winter of capital, as well as the inherent disadvantage of lack of algorithms. Something worth thinking about. It's easy to think of last year's VR industry. Now the industry's hype about artificial intelligence is very similar to the VR's collective bullishness at the beginning of last year. The environment and current situation of artificial intelligence are also the same as that. VR is exactly the same, the same capital investment, blooming everywhere, the same future technology nature, the same high expectations of the market, but now semi-dead VR / AR industry is likely to become a precursor to artificial intelligence companies. For the artificial intelligence that has become the outlet, we should maintain enough rationality in the virtual fire and boom. Compared with public opinion's rally, in fact, investment in the area of ​​wages and wages has apparently begun to become more cautious. This also confirms the fact that capital Wintering has said, and this trend may take a long time with the progress of the Internet economy. As can be seen from the figure, the investment quota has significantly improved since the end of 2014. This is consistent with the emerging time of the start-up companies. Even though the quarterly fluctuations are large, the total annual amount is indeed rising. However, from the point of view of the financing stage, since 2005, the capital has been mainly concentrated in the seed round, and the proportion of A, B, and C stages has been reduced as a whole. Moreover, there are statistics showing that over the five-year period from 2011 to 2015, more than 65% of the global financing occurred in the seed/angel round or round A, and there were only 20 rounds of financing in the D round and beyond. This shows that although artificial intelligence is very hot, investment is becoming more and more rational. If most artificial intelligence companies do not have clear practical values ​​or technical products, it will be difficult to impress investors' hearts with confidence. Their future financing prospects are indeed Not optimistic. So telling stories and engaging in marketing is not really an AI process. Even technology giants in this field have not been able to create an artificial intelligence product with perfect performance and large user scale. Most of the products of artificial intelligence can only be regarded as artificial artificial intelligence. Instead of investing too much in the marketing of products, it is better to pay attention to some vertical subdivision areas that the giants have not touched, and to seek opportunities from them. It takes time for artificial intelligence technology to become commercial value The application of artificial intelligence involves both dedicated applications and general-purpose applications. The former covers most areas of current domestic artificial intelligence applications, including face and speech recognition and service robots, while the latter focuses on finance, medical care, and intelligence. General solutions for home and other areas. No matter what the application of the field is, it has a clear business model. For example, today's headlines that rely on artificial intelligence and black box algorithms have risen strongly and passed through suitable content to the right people to accumulate massive users; Apple's Siri, Amazon's Alexa and other voices. Interactive technology successfully integrated with hardware products and earned a lot of profit for it. However, this also illustrates another grim fact: At present, the more mature artificial intelligence products in the industry come from the Internet giant. Under such pressure, start-ups often use pseudo-intelligent products as the capital for storytelling in order to quickly obtain investment or earn eyeballs. This phenomenon is extremely common, and it also creates a situation in which "hot demand for products is cold" on the market. This is very similar to the state of the VR market in the last two years. Numerous manufacturers have launched their own VR hardware or software programs with great popularity. However, these low-performing cheap goods not only failed to take advantage of it, but also destroyed most of their consumption. Person's willingness to buy. This is also true in the field of artificial intelligence. As Zhou Hongyu said, “Today, come out and be a company, if you don’t say that you are using deep learning, artificial intelligence, you are embarrassed to mix”, “Each pig basically has an artificial intelligence on its body. The tag". A large number of pseudo-intelligent products simply add ordinary electronic products to the networking or search function and can be touted in the name of big data and deep learning, such as some wearable devices and smart set-top boxes, and even more. A cup with a digital thermometer will dare to call smart. The existence of these products has caused the misunderstanding of artificial intelligence by ordinary users to a great extent, and it will also have a negative impact on the application of the industry in the future. And even if there is a technological breakthrough, it will take a long time to turn from technology into great commercial value. Since the birth of the concept, artificial intelligence has been in existence for 60 years. Even during these years of scientific and technological explosions, there are no phenomenal products that have swept the market. Each branch is also facing its own know-how. Even if you break through the know-how, it does not mean that this product can be successful. Wang Xiaochuan said: If there is no technical content at the beginning, it will be very difficult to become a product; further, even if there is technology, it is not necessarily a product, not necessarily the size of the user; and further, even if there is technology, there is a user Scale does not necessarily bring great business value. All in all, the moment when artificial intelligence requires technology to be developed, it needs more patience to wait for the balance of time, place, and people. So, is artificial intelligence now available to benefit people? It is the most reasonable and healthy growth route to drive the market outbreak with technology maturity, just as the smart phone has a strong spike function. In particular, artificial intelligence products with strong sci-fi colors should be gradually and step-by-step. They must be regarded as victims of the wave of science and technology because of the imaginary bubbles generated during the business upsurge. VR has been played badly last year, and it takes a while to recuperate to have a chance to re-emerge. I hope AI's luck will not be so bad.