This time, it’s really not far from the birth of “AI First Unit”.
Just a few days ago, a document from the Hong Kong Stock Exchange disclosed that SenseTime was expected to be listed on the Hong Kong Stock Exchange soon after a listing hearing. The IPO plans to raise US$1 billion and become the world’s largest AI unicorn IPO. Previously, Megvii Technology and Yuncong Technology also held meetings on the Science and Technology Innovation Board in the second half of this year. So far, three of the “AI Four Little Dragons” have obtained the “Customs Clearance Certificate” in the secondary market.
Regarding this wave of listings, it seems to the outside world that it is an inevitable result. The artificial intelligence industry has a long investment return cycle, high investment, and serious losses. It has become a consensus, and the investment market is no longer blindly betting on the AI track as it did in previous years.
According to the prospectus, Shangtang’s team is mostly scientific research groups: 40 professors and 2/3 of its 5000 employees are scientists and engineers.
“Everyone can’t wait, can’t wait any longer.” An investor bluntly said to the investor. “When effective hematopoiesis was not achieved, going public became the common and only choice for these unicorns.” “(For AI companies) the current situation is wherever they can go.”
On July 29 last year, Megvii CEO Yin Qi said that all AI companies have entered the valley of life and death. He believes that the AI industry has entered the deep water zone, and the travel time may be in the next 18 to 24 months.
This makes some AI entrepreneurs also have a normal anxiety: how to promote the integration of AI technology and industry more quickly, how to get out of the bubble period, and how to not be eliminated.
Until today, have the AI players represented by the “Four Little Dragons” stepped out of the “Valley of Survival”? Maybe soon, maybe it will take longer.
Note: The content of this article mainly comes from interviews by Pencil Road reporters and public information from authoritative sources. The arguments are inevitably biased and there is no deliberate misleading.
The “AI” four little dragons compete to be listed
On the evening of November 22, according to documents from the Hong Kong Stock Exchange, SenseTime passed the Hong Kong Stock Exchange listing hearing. The joint sponsors are CICC, Haitong International and HSBC . It is expected to be listed on the Hong Kong Stock Exchange soon. Capital of 1 billion U.S. dollars.
It is worth noting that SenseTime completed the last round of financing in June 2021, with a post-investment valuation of over 12 billion U.S. dollars. The capital market recently predicted its valuation to reach 13 billion U.S. dollars. If it is successfully listed, SenseTime may become the world’s largest IPO in the field of artificial intelligence.
In fact, Reuters reported that SenseTime’s Hong Kong stock IPO was approved. SenseTime’s external response at the time was: “Thank you for your attention, the company does not comment on this.” Now, with the official announcement of the Hong Kong Stock Exchange, the truth has been revealed.
Established in 2014, SenseTime is an AI software company focusing on computer vision and deep learning technology. Its main business includes four sectors: smart business, smart city, smart life, and smart cars. Among them, smart business and smart city have taken up SenseTime’s revenue, and their proportion has increased year by year. In the first half of this year, the two major businesses accounted for more than 80% of the total revenue.
According to the prospectus, SenseTime is the largest company in Asia in terms of revenue in the AI industry. In 2018, 2019 and 2020, Shangtang’s revenue was 1.85 billion yuan, 3.03 billion yuan, and 3.45 billion yuan, respectively.
However, while the revenue performance is improving, the loss of SenseTime’s technology enterprises is also worth paying attention to. In 2018, 2019, 2020 and as of June 30, 2021, SenseTime had net losses of 3.433 billion yuan, 4.968 billion yuan, 12.158 billion yuan and 3.713 billion yuan, respectively.
As an AI unicorn, SenseTime has always been favored by capital. According to the prospectus, SenseTime has completed 12 rounds of financing, with a total financing amount of US$5.2 billion (approximately RMB 33.658 billion). In terms of shareholders, Shangtang’s largest shareholder is still the founder, Professor Tang Xiaoou, who holds 21.73% of the shares. Among the external shareholders, SoftBank holds 14.88%, Alibaba holds 7.59% of Taobao China, Chunhua Capital holds 3.08%, Silver Lake Capital holds 3.05%, and IDG Capital holds 1.42%.
As for the reason for being sought after by capital, SenseTime’s co-founder Xu Bing once disclosed a part in an interview with the media: “All of our investors are currently getting returns far exceeding their average rate of return. Silver Lake Capital is a well-known investor in the market. One of the investment institutions in the TMT industry, the average return on investment in the past 30 years is 37%. This is a very terrible figure, and it should be one of the best in the global TMT investment field. Investors including Silver Lake, after investing in us , The rate of return is much higher than this figure. We have allowed all investors to make a lot of money.”
Forgot since when, the “AI four little dragons” spread in the arena. From financing to listing, the pace of development of the “Four Little Dragons” is highly coordinated. Megvii, the first of the “Four Little Dragons” to apply for listing, submitted a prospectus to the Hong Kong stock market in 2019, but terminated the listing process in Hong Kong within ten months. The meeting will be held in September this year. In addition, Yuncong submitted a prospectus to the Science and Technology Innovation Board in 2020, and it has also successfully passed the meeting in July this year. At present, the only one that is a little behind is Yitu. After its application for listing in 2020 was accepted by the Shanghai Stock Exchange, it took the initiative to withdraw its application for listing in June this year.
Why did the “Four Little Dragons” race to market? Investor Yang Jun (pseudonym) analyzed the Pencil Road and said: “Everyone can’t wait, can’t wait.”
The high investment in the artificial intelligence industry, long return periods, and serious losses have gradually become common knowledge, and the investment market has no longer blindly bet on the AI track as it did in previous years. “For this reason, when effective hematopoiesis is not achieved, listing has become the common and only choice for these unicorns.”
Regarding the listing process, some investors sighed: “(For AI companies) the current situation is wherever it can go.”
Invisible AI, visible loss
Worrying about money is the norm for AI entrepreneurs and has nothing to do with the size of the company.
Opening the prospectus of the “AI Four Little Dragons”, the common point is the failure to achieve profitability. According to rough statistics, the “Four Little Dragons” accumulated losses of over 40 billion in three years. Although losses include “fair value losses” (mainly losses caused by employee options), after deducting non-recurring gains and losses, the net adjusted losses of all companies have decreased, but the numbers are still huge.
Take this upcoming Shang For example, in the earnings report, the Shang Dynasty loss due to the huge R & D expenditure, which is the research payroll costs accounted for the bulk of developers.
SenseTime Lianchuang and CEO Xu Li once said, “The battle of artificial intelligence is the battle of talents.” Therefore, since the completion of the angel round of financing in 2014, SenseTime has begun to recruit talents in a carpet style. The prospectus shows that as of June 30, 2021, SenseTime has 5,286 employees, of which 3,593 are R&D personnel (more than the sum of the other three R&D personnel), and 40 professors are leading the R&D work.
Therefore, there are also views that: “AI companies are a carnival for scientists, but academic papers cannot generate revenue. Someone has to pay for the high salaries, expensive clusters, and electricity bills of employees.”
On the other side of huge losses, investors are becoming more and more eager to make profits. Huasheng Securities believes that the current AI startup companies have grown for many years and have shifted from “technical breakthroughs” to “commercial growth” cycles, and the choice of capital for them has also shifted from technological breakthroughs to the stage of commercialization.
However, the commercialization of AI is not easy. The founder of an AI company previously shared his feelings with Pencil Road: “AI entrepreneurship, death will follow.” The core reason lies in the high investment in technology, but the difficulty of making money.
From the current point of view, the commercialization of AI is relatively concentrated, and the “AI four dragons” are all competing in the security and financial fields. The general direction of “same results with the same work” and the overlap of some vertical businesses have made competitors move from avoiding homogeneity to homogeneity.
Of course, AI’s difficulty in making money is caused by many factors. An industry analysis, “Four Little Dragons”, though they have a very powerful technical strength, but in the market to attract customers, often relatively simple mass products, it is difficult to achieve very high profits.
In other words, although AI technology is rapidly iterating in the high-precision direction, the demand on the market is still relatively low-end. “For example, Shangtang’s difficulties stem from the huge gap between vision and reality. The current AI technology cannot achieve a platform-based business model, and fragmented customized scenarios cannot support excessively high valuations.”
An employee of one of the “Four Little Dragons” said that the fact that the “Four Little Dragons” has not yet made a profit is mainly due to the characteristics of the industry itself and technical limitations.
In terms of model, most domestic enterprises (especially most small and medium-sized enterprises) have not formed the habit of paying for services, so most of the AI company’s business can only be to G or to major customers. However, to G and to major customers usually have a high degree of project customization, poor project reusability, low marginal effects, and excessively high costs.
At the technical level, the current AI technology is not mature enough, and only recognition (mainly face recognition) technology can be implemented, which leads to a relatively single product and service and a relatively small market. The market has many demands for AI technology, but the accuracy of many demands cannot be commercially available for the time being, and some demands cannot even be fulfilled.
In summary, the ability to burn money is amazing and the progress of commercialization is slow. This is part of the reason why AI companies, including the “four little dragons”, have been questioned by the outside world.
For investors, the previous valuation level is too high, but the development momentum of related companies currently seems difficult to maintain high growth, and there is no stage of profitability.
More importantly, it is not only the loss at this stage, but also the probability that AI companies will continue to lose money in the next few years.
When the tide goes away
Those issues that were once covered by vents and concepts are exposed to the AI industry naked as the capital ebbs.
In the face of the dilemma, the “AI Four Little Dragons” not only seek new capital from the secondary market, but also need new stories. The “Four Little Dragons” are also seeking transformation, beginning to seek differentiation in business models in more vertical fields.
In the prospectus, SenseTime stated that 60% of the IPO funds raised will be used for the company’s R&D investment; 15% will be used for business development, mainly emerging business opportunities and domestic and foreign market penetration; 15% will be used for investment and mergers and acquisitions, SenseTime Said that it will seek foreign strategic investment and mergers and acquisitions opportunities to expand the AI ecosystem and its influence in all walks of life.
In fact, SenseTime has been continuously expanding its business scope, and gradually expanded its front line to autonomous driving, smart phones, medical and health, entertainment, advertising, new retail and other fields. After that, it announced its entry into AI education and established its sub-brand SenseTime. educate.
When the concept of Metaverse was born, the outside world gave AI companies ideas, suggesting that they associate with Metaverse as quickly as possible.
Indeed, the realization of the meta-universe is inseparable from AI technology. SenseTime’s prospectus also mentioned Metaverse 50 times, and stated that the company has built a multi-layer infrastructure for empowering IOT devices and driving Metaverse. Its SenseMARS software platform has become one of the largest Metaverse enabling platforms. Contains a new meta-universe experience supported by more than 3,500 AI models. The company has been developing meta-universe-related technologies that can be widely deployed on cloud and IoT devices. It is expected to build a ubiquitous entrance to the meta-universe and create a truly immersive experience. Meta universe experience.
However, at present, Metaverse is still in the conceptual stage. In addition to telling stories, it is still uncertain how much help it can bring to AI companies. What the capital market wants to hear is no longer a future-oriented “sci-fi” story, but actual commercial content.
When Tang Xiaoou and SenseTime’s CEO Xu Li founded the company, they had a consensus: “Don’t do superior technology, but transform technology into business and productivity.” However, how to become a profitable commercial company today? For this question, SenseTime has not yet handed in a satisfactory answer to the outside world.
This problem is also true for other AI players. “All AI companies have entered the’valley of death’, and everyone has high expectations. When expectations fall, it is difficult for everyone to predict which AI company can cross this’valley of death’. “last July 29, Kuang as the technology CEO printed in odd years in the media said so on the experience. He believes that in the next 18 to 24 months of the bubble period, AI companies that do not bring true value will be eliminated. According to him, the current AI companies are still in the bottom.
IPO can continue the lives of the “AI Four Little Dragons”, but it cannot solve all problems. Listing is not the end, but the starting point. When the halo fades, how to live better and better? This has become a problem facing all AI companies at the moment.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/a-group-of-scientists-made-an-ipo-in-7-years-12-billion-us-dollars/ Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.