Wind 2018 | When the block chain becomes a new outlet, will AI start-up financing open hard mode?

Q: The 2017 AI industry's biggest progress is what.

A: The valuation of a start-up company.

Behind this industry joke is the 2017 China AI sector investment and financing "prosperity".

But 2018 years on, such a semblance of prosperity, or will be difficult to reproduce. Ai VC is ushering in a rational return. And more AI startups may welcome the hard model of financing.

(Note: The data from the whale, the investment heat is mainly based on the behavior of investors on the whale, can reflect investors on the project, track attention)

In the last week of December 2017, when the first-tier market launched a whale, the investment heat of AI circuits began to fall in line, with investment data falling from 92 to 73. After experiencing a rapid growth, smooth upgrade, a sharp drop.

This phenomenon may not be in the expectation of ordinary readers. After all, from the perspective of the 2017, especially in the fourth quarter, the AI field of investment and financing is particularly lively, the track not only suction and gold absorption.

October 2017 AI education company 乂 Angel Round financing up to 270 million yuan, this money is likely to invest in Ofo's B-round; AI chip manufacturer Cambrian A is the unicorn, before the Chinese company to achieve this result is Ali Sports, melon seeds used cars, these companies often have a large platform resources injected , a humanoid robot company is expected to have a valuation of 4 billion U.S. dollars, computer vision Company's soup technology was spread to 3 billion U.S. dollars in valuation financing, two companies in user data, financial data may not be prominent, and before the racer reached the 3 billion U.S. dollar value of the month of life has reached 150 million.

But dismantling such a prosperous appearance, AI investment cooling and cooling, funds, resources focused on the first echelon, in 2017 presented a clear precursor.

AI investment Cooling, return to rational

(Note: Data from whales, 2017 data for 2017 1-November data)

(Note: Data from whales, 2017 data for 2017 1-November data)

2017 1-November, China's artificial intelligence field first-level market investment and financing transactions amounted to 18.94 billion yuan, accounting for the first-tier market investment and financing transactions of 2.38%, the number of transactions reached 219, accounting for the total trading time of 4.892%, a slight decline compared to the 2016 highs.

(Note: Data from whales, 2017 data for 2017 1-November data)

From the average amount of financing, 2017 is basically the same as 2016, single financing is around 86.5 million yuan, did not present the transaction amount, the number of transactions close to the downward trend.

Second, funds, resources focus on the big market, the first echelon

Actually cold, but the appearance of fanaticism, with funds, resources focus on the first echelon. The Matthew Effect makes the first echelon get a lot of external attention.

We have 2017 years of artificial intelligence in the field of the highest amount of 5 financing, 5 of the volume of financing has exceeded 8 billion yuan, accounting for artificial intelligence in the field of the total financing of about 45%. In fact, even in the top 5, the gap is also very clear, Top1 see the technology of the C round of financing is Top3 out to ask the D-round financing amount of more than 3 times times. 2017, the science and technology, commercial soup technology two each completed a round of more than 400 million dollars of financing, accounting for all the artificial intelligence investment in the field of about 30%.

This Matthew effect was particularly pronounced in 2017. With the technology, Business Tang Science and technology in the field of computer vision face recognition as an example, the track of the company up to more than 100, but 2017, Kuang-tech, commercial soup technology two financing more than 870 million U.S. dollars, accounting for the track financing amount of 65%. 36 Krypton learned that the July 2017 commercial soup to complete the 410 million dollar round of financing, the financing disclosed a period of time, investors on the track in the financing of the non-first echelon of the company's attitude is clearly cautious, the second half of the circuit on the track a round of financing very few. This is a long time, the business Tang technology of this round of financing, a total of 20 institutions involved, these investors may bring about the relevant customer resources.

In the smart voice direction, the most competitive in the current speech recognition direction, because Baidu, Microsoft, the Hkust and other large companies to participate in the strong, there is an earlier venture out of the door to ask, think must have formed the first advantage, direct auction companies to obtain financing probability is low, access to financing companies are more from the upstream and downstream, such as the upper reaches of the acoustic device , microphone array, downstream natural language processing, industry applications.

The reason for this phenomenon, we think that with the AI direction entrepreneurial characteristics. National AI Entrepreneurship is a data resource, talent technology driver. New technology start-up at this stage also mainly from the technology, the application of the market to G, to B-oriented, often with the macro-body scene, business combination. The direct business data is one of the important competitive barriers, and the successful customer case is the important thrust to enlarge the business. So the Matthew effect will be obvious.

On the other hand, the first echelon of the company's active strategic and strengthened the Matthew effect. With many areas of exclusion, AI start-ups are more inclined to "hug the thigh", accelerate the actual landing, the first to obtain industry data and experience. In the "AI era of strategic investment: from for financial investment, to be the financial investment," the article we have mentioned that the number of strategic investment in 2017 increased, and the strategic investment phase earlier, even the financial investors began to take the company strategic investment as one of the criteria for investment.

In fact, many companies are actively in the company, interested in wooing strategic investors and national team investors. such as in AI chip circuit, almost well-known companies behind the semiconductor manufacturers, the Internet large companies figure. Yiu Songzeng, chief executive of Science Technology, said in a 36-krypton interview that he had made concessions in valuations in an effort to secure an investment in Ant's Gold suit. While the science and technology of the C-round financing, the official news campaign has also actively emphasized the status of the national team of investors.

In addition to focus on the first echelon of funds, but also gradually focus on the big market. 2017, the industry's view of artificial intelligence gradually unified, artificial intelligence is to enhance the efficiency of the industry tools, tools need to be combined with the scene, business.

In theory, artificial intelligence can replace or partially replace repetitive, mechanically strong work. But whether AI companies can reap economic benefits depends on their ability to pay, and how much the AI company continues to earn depends on the size of the industry. Therefore, the technology can be applied to high customer price high margin of the large market companies, more favored. Access to financing, especially large-sum financing or higher valuations of AI companies, the basic security, automotive, finance, Internet of Things as the main object of service.

Third, the start-up company or to meet the financing hard model

When funds, resources to the first echelon of the company focus on the first to second echelon of the start-up companies, or will usher in financing hard model.

Although early investment is to admit the low success rate, Bo High-yield projects, but the current AI field of good projects more and more difficult to find, more and more expensive, success rate is important. Under this mindset, the first echelon is likely to get a higher premium.

Since the second half of 2016, the overvalued bubble has been questioned around the industry. Up to now, the artificial intelligence field few has the profit of a round after company, the well-known company is likely at the earliest 2018 years only then may realize the year profit and loss balance.

Whether there is a bubble in the debate behind the subtext is often recognized by the industry and the market, but can not accept such a price. In the short term, however, the scarcity of talent, high human costs and low valuations are unrealistic. In the field of artificial intelligence, entrepreneurship is technology-driven and talent-driven. But in a certain period of time, talent and technology are often stable supply. The domestic AI talent mainly comes from the American and European Research Institute, the internet big company, in the past two years the domestic AI start-up company basically consumes the most talented person. And to the younger and outside the United States and Europe, such as commercial soup began in high school to scramble for talent, according to the figure in Singapore set up an office.

(Note: Above is the salary level of a headhunter, the picture is from the new intelligence Yuan)

In addition, the current domestic artificial intelligence in the field of high salaries, start-up companies operating costs relatively high. According to market feedback, "Master's degree in 3 years of machine learning algorithm expert salary on the general can reach the annual 60.8 million, while the master graduated 8 years/PhD graduated 5 years, if the development to the position of Algorithmic director, pay more than 1.5 million". In the short term, industry average wages are hard to reduce.

In addition, in the past because AI project valuations are too high, often behind the start-up companies have gathered a number of investment institutions, many well-known start-ups of the founding team held shares have been diluted or is about to be diluted to 50% below, and to allow a number of institutions to accept undervaluation, coordination may be relatively complex.

Based on this, we believe that most startups are less likely to lower valuations in the short term. Therefore, 2018 capital may reassess the market already has the target, each domain 12th Echelon company obtains the financing probability to be increased, but does not head company obtains the head organization the difficulty to be able to continue to increase, even obtains the financing the difficulty also to be promoted. In addition, it is worth noting that the general start-up in the face of the trough, will often produce brain drain, this reverse will affect the recognition of the company's investors.

Late-stage projects are too expensive, the company does not occupy a sufficient share in the market, it is likely to make new entrepreneurial projects to obtain capital favor. Early project premiums may be a new attraction for talented people to start a business. But the past few years have consumed academics, big companies, quality talent, the market is too few quality new projects, it is likely to make early investment institutions to save entrepreneurial team. The new entrepreneur is likely to come from a developing start-up company. Given the difficulty of landing new technology companies, the direction of entrepreneurship is likely to be the same, similar or close to the original company.

As a result, funding is tilted towards the head and new companies, or it will become the norm in the 2018, and it is a problem that early startups need to deal with.

Iv. funding or flow to other areas

Perhaps even more worrying than criticizing the industry is that there are no bubbles in the industry at first.

In the age of the rise and the new technology, the song of Ice and Fire, we have counted the average single amount of funding in each field. 2017 1-November, the average amount of single financing in the field of artificial intelligence amounted to about 86.5 million yuan, only about half of the average financing amount (about 170 million yuan) in the whole field. After statistics of the whole field of investment data, we found that the logistics, electricity, medical, education and other areas of the average amount of single financing, 2017 increase of more than 300%, the tuyere is changing.

Many domestic investment institutions, artificial intelligence are subordinate to the High-tech direction, it is likely to include robots, drones, IoT, semiconductors, block chain, AR, VR, pharmaceutical, medical equipment, 3D printing, new energy, materials, commercial aerospace and many other subdivision of the track. And from the current institutions have been out of the sharing of content and 36 of the private understanding of Krypton, many organizations are also looking at the block chain, pharmaceutical and other new directions. And from the current heat of opinion, block chain has become no doubt the new technology tuyere.

This is clearly not a particularly good news for AI startups. To get more money earlier may be the best solution.