Shulou(Shulou.com)06/02 Report--

Why is AI always used to kill chickens?

Core points:

Capital regards artificial intelligence as the key to the next era, and no one wants to miss the opportunity of the times, so there are frequent large amounts of financing in the field of artificial intelligence, resulting in bubbles. Artificial intelligence technology falls to the ground slowly, the scene is narrow, and commercialization takes time, and there is uncertainty about the future of related enterprises. The real value of artificial intelligence technology lies in solving the pain points of the industry, not the sci-fi scenes depicted by all kinds of grandiose marketing, and requires companies to keep their feet on the ground rather than telling stories.

Nowadays, from the old to the small, there is almost no one who does not recognize artificial intelligence, although there may be some differences between these understandings.

For their elders, they may still need help from their children in online shopping, but widespread media reports have instilled in them the idea that "AI is about to replace human work"; for children, various "AI programming lessons" are for them. Although parents themselves may not understand AI, no one wants their children to lose at the starting line of a new era.

The promotion of capital is the key reason why the media and businesses are so enthusiastic about artificial intelligence.

According to data from IT Orange, investment and financing events in the field of artificial intelligence have increased significantly since 2016, and large financing events still occur frequently in 2018 and 2019, which are known as the "cold winter of capital".

In the capital boom, a number of start-ups have emerged in the artificial intelligence track, including many high-profile unicorns, such as Cambrian, an AI chip company founded in 2016. In March this year, the Cambrian officially knocked on the gate of Chuangban, and the pre-IPO valuation has reached 22.2 billion yuan.

Of course, the giant BAT will not let go of the tuyere, and it has not stopped investing in the field of artificial intelligence while doing research and development in person. This past April, Tencent invested in two companies in the field of artificial intelligence: Zhuo Vision Zhitong and Light Vision.

Over the years, tuyere has appeared one after another, some are just a gust of wind (live answer), some are mired in a quagmire of losses (sharing economy), but artificial intelligence is still hot, and the difference is largely due to people's different expectations of artificial intelligence. The mainstream view in public opinion is that artificial intelligence is an important force to promote the fourth industrial revolution, which is equivalent to steam engine, internal combustion engine and Internet.

In contrast to the popularity of capital, artificial intelligence still does not seem to have a killer application, and science fiction scenes are a long way off, so the "bubble theory" comes out from time to time. Many people began to ask: is artificial intelligence real fire or false fire?

AI gale continues to be strong

In China, the outbreak of artificial intelligence in public opinion and capital markets was in 2016, marked by the "man-machine war" between AlphaGo and Lee se-dol.

In March 2016, AlphaGo beat world champion and professional nine-stage go player Lee se-dol at 4:1, then faced dozens of go masters online without losing 60 games. In 2017, AlphaGo beat world No. 1 go champion Ke Jie at 3:0.

AlphaGo vs. Lee se-dol

The great victory of AlphaGo made the world fully aware that the chess power of machines has reached a height beyond the reach of the human mind, and there are endless comments that the technological revolution will subvert human society. Since then, domestic giants have acted frequently in the field of artificial intelligence:

In July 2017, Baidu shares soared when the then vice chairman of the board of directors, group president and COOLuqi proposed "All in AI" at Baidu AI developers' conference. In October 2017, Alibaba announced that he would invest hundreds of billions to set up Alidamo House to join the battle in the field of artificial intelligence. Tencent has established Tencent AI Lab, Robotics X, and launched the Tencent AI open platform.

Foreign Internet giants are enthusiastic about artificial intelligence earlier. For example, Zuckerberg put forward Facebook's "10-year plan" in early 2016, which focuses the company's future strategy on new technologies such as artificial intelligence (AI), virtual reality (VR) and augmented reality (AR). Zuckerberg has also had several "war of words" with Musk on the Internet because of his different views on artificial intelligence.

Both Silicon Valley giants and domestic BAT regard artificial intelligence as the key to the next era, and it is not difficult to understand the logic that they attach so much importance to the technological revolution: these enterprises have grown up at the "top of the wave" of the Internet and have a clear understanding of the power of technological change. Once they miss the technological revolution, what awaits them may be the fate of falling from the peak.

Capital is also aware of the power of the technological revolution. PC and the mobile Internet have made BAT, TMD and other companies. The development of artificial intelligence may give rise to new exponential growth opportunities.

As a result, there have been frequent large amounts of financing in the field of artificial intelligence, such as the completion of D-round 's $750 million financing in May 2019 and the $1 billion D-round financing from Softbank Corp. in 2018.

ZhenFund founder Xu Xiaoping, a famous supporter in the field of artificial intelligence, once said that "there is no doubt that heavyweight companies will be born in the field of artificial intelligence", "if it was Internet + a few years ago, then it must be AI+ in the next five years" and "not enough money has been invested in AI." According to data from IT Orange, ZhenFund has made 90 investments under artificial intelligence, making him the most active investment institution in the field.

As for the impact of artificial intelligence on human society, Kai-fu Lee, founder of Innovation works, pointed out in his book "the Future of AI" that nearly 50% of human work in the future will be replaced by artificial intelligence within 15 years. Shen Nanpeng, global managing partner of Sequoia Capital, believes that artificial intelligence will bring another great liberation of human hands and brain, and the meaning of work will really change from "labor" to "creation".

Generally speaking, "artificial intelligence will have a profound impact on human society" has become a consensus. However, judging from the current actual landing situation, there is still a big gap between the ideal and the reality.

The market is mixed, and the dilemma of commercialization is difficult to solve.

Thanks to the enthusiasm of capital and the media, AI has become the most fashionable concept these years. Both traditional industries and new economy companies are keen to label their products as "AI".

In sharp contrast to the various fantasies of the giant founders or investors on AI, smart scenes often only exist in the enterprise's PPT, and in reality, the performance of various "AI" products is often funny, which is far from the ideal "AI intelligent butler".

Take the AIoT and Smart homes, which are widely distributed by many manufacturers, as an example. Apart from answering "what's the weather like today" and playing Guo Degang's crosstalk, most intelligent devices cannot understand even slightly more complex instructions.

In the AI craze, manufacturers often want to use technology to endow every part of life, but due to the current actual technological breakthroughs of AI, its landing application often appears to have some chicken ribs, such as opening and closing curtains with voice and voice interaction with furniture. It is not so much that AI brings convenience to people's life, but people are willing to "add trouble" to practice and upgrade in order to improve the ability of AI.

The "empowering" of manufacturers is quite a bit like "holding a hammer in hand, everything is a nail". Instead of bringing convenience, technology has become an "addition" to life.

As for the idea that AI will replace human work, there is no sign yet.

Many companies claim that customer service robots with AI blessing can effectively solve users' problems, but people who have tried them all know that the answers given by the machines are often routine actions that are set up and can only be solved manually if they encounter real complexities. In terms of writing, AI has indeed been able to generate articles, but such articles are often in a fixed format, and the process of generating them is more like filling in blanks and permutation. The AI writing robot is currently powerless for content that needs to be creative and communicate with people.

Compared to the unsatisfactory performance of AI, it is more common in the market that AI has completely become a marketing concept, and the product actually uses only very basic instructions, or even a complete hoax.

Kaifu Lee publicly complained about the grandiose atmosphere in the field of artificial intelligence in 2017: "recently I met a lingerie maker who also said that he was an artificial intelligence enterprise, which is a very abnormal phenomenon." now, the bubble in the field of artificial intelligence is particularly serious. "

In March last year, venture capital firm MMC reported that about 40 per cent of European start-ups were fake AI companies and that there was no evidence that they used artificial intelligence, but with the AI tag, companies could attract 15 per cent more money.

A typical example is Engineer.ai in India, which claims that its AI can customize APP and websites for companies that lack engineers or don't know how to program, that is, it doesn't matter if they don't understand technology, just give it to AI if you have an idea.

Engineer.ai painted a beautiful picture, successfully attracting investment from DeepCore, a special artificial intelligence fund owned by SoftBank, but it was exposed by foreign media last year that the so-called "AI" used by Engineer.ai does not exist, and the code is actually written by human programmers.

In fact, it is not uncommon to use a real person to pretend to be AI.

In the field of new economy, whenever there is a new concept can always attract a lot of hot money, sharing economy, VR/AR and other concepts are hot when such a situation has occurred. After all, there are only a few head companies, and the influx of capital will naturally flow to companies that do not have hard power and only rely on storytelling to circle money, resulting in a bubble.

Bubbles in hot areas are not uncommon, and many investors know that artificial intelligence companies are generally overvalued, but no one wants to miss the next era because of caution, but interestingly, even the leading companies in AI don't seem to be able to digest the enthusiasm of capital.

Due to too much financing, "VC of AI" has become a unique landscape in the field of artificial intelligence. Leading companies such as Shangtang Technology and Xanshi Technology have set up their own war investment departments and directly put projects out of the market; Sibichi, Zhongke Shituo, Aibee and others have jointly set up industrial funds with traditional VC and PE.

The reason why we want to burn other people's money while taking other people's money to burn, the essence is that AI's actual business can not support the valuation.

AI technology needs a long period from research and development to landing, and many companies will promote the breakthroughs of their own AI technology in the academic field, but it is not easy for technology to move from laboratory to commercial application. Some numerical breakthroughs in research papers can shock the academic circles, but may have little impact on the actual experience of products. In addition, the current AI landing scenarios tend to focus on some specific areas, the application scene is narrow, can not quickly get high returns.

From the point of view of the business model, the head AI company is mainly To B business, while To B is a slow business, so it is impossible to rely on capital to "feed" the business as quickly as the To C business such as bike sharing.

The landing is slow, the scene is narrow, and commercialization takes time. AI, which has taken a large amount of financing, can only rely on investing in other companies to prop up its valuation and respond to capital expectations.

As the dilemma of the commercialization of artificial intelligence is difficult to solve, the capital has gradually become cautious in the past two years, and the investment logic in the field of artificial intelligence has also been iterated. The focus of investors has shifted from the technical ability of the company team and the number of papers to "how does your AI land?" "AI landing" replaced "AI worship".

Prick the bubble, where is the way forward?

Judging from the current situation in the capital market, the strong wind of artificial intelligence is still strong, but compared with the previous calls in the public opinion field, the seriousness of the concept of AI has been dispelled to a large extent, and the arguments about "artificial intellectual disabilities" and AI bubbles appear frequently.

The emergence of AI bubble theory is largely related to the grandiose marketing of manufacturers. When the promotional materials of all kinds of products are full of "AI", but few of them really feel the role of technology, it is hard not to feel that this is a fake tuyere.

Looking back on the course of the technological revolution, pseudo-tuyere and bubble theory also appeared in Internet technology, but finally got a correct name. In order to return to the value of artificial intelligence, the first two technological revolutions (computer software technology and Internet technology) are effective references.

In the era of computer software, people hand over the previous manual work to the computer, that is, to realize the "informationization" of work and behavior. Informatization enables data to be produced in large quantities, which is equivalent to human activities being recorded in the form of data. At this stage, computer software has improved the efficiency of human activities, the rise of Microsoft, Oracle and other well-known software companies.

After a large amount of data was produced in the era of computer software, the emergence of Internet technology enabled the flow of these data. With the commercialization of the Internet and the widespread flow of data around the world, the "traffic economy" arises. Both e-commerce on Taobao and social networking on QQ are essentially business models based on mobile data. With the development of Internet and communication technology, the amount of data is becoming larger and larger, and new technologies such as "cloud computing" emerge as the times require.

With huge and mobile data, how to develop the deep value of these data determines whether the era can enter the next stage. The value of artificial intelligence technology lies in the analysis and correlation of massive data, and then the formation of knowledge that can guide decision-making, which is the key to solve the pain points of the industry.

The pain points faced by different industries are different, such as the financial industry is facing cost pressure, single products and services, transaction fraud, etc.; the problem of uneven distribution of resources in the medical and educational industries is prominent; it is difficult for the retail industry to grasp customer demand, advertising efficiency and so on. For the industry, the combination of artificial intelligence and scenario means reducing cost and increasing efficiency, realizing transformation and upgrading.

At present, the government and enterprises are combining artificial intelligence technology with industrial scenarios. At the government level, artificial intelligence is an important one of the seven sections of "New Infrastructure"; at the enterprise level, the giant's AI layout is also closely related to the industry.

In other words, artificial intelligence has a broad demand, and its business model is to infiltrate various industries and improve the efficiency of the industry. This process requires time and continuous investment, but it is also a productivity iterative trend.

The landing of artificial intelligence may not be as "sexy" as you think, but the impact on life is only increasing. Forget the grandiose marketing and Iron Man's Jarvis, those AI who create demand out of thin air and add to life will eventually disappear in the long river of time.

Http://www.woshipm.com/ai/3795579.html

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.