AI “Four Little Dragons”, the death of unbearable cash flow

The industry is not big enough, the development of the enterprise itself is restricted, and the industry is big enough, and it is easy to attract giants.

AI "Four Little Dragons", the death of unbearable cash flow

On June 23, 2017, Wuzhen, Zhejiang, Google’s artificial intelligence Alpha Dog played a game with the world’s number one talented chess player Ke Jie. Four hours later, Ke Jie was defeated.

With the appearance of Alpha Dog, people realized that artificial intelligence is already so smart.

2017 is also considered to be the first year of artificial intelligence, and the AI ​​track has begun to become popular. Industry representatives Shang Dynasty Technology , Kuang depending on science and technology, according to map technology , cloud from science and technology (collectively referred to as “AI Tigers”) also received another round of financing.

Due to the difficulty of commercialization, the profitability of the AI ​​four dragons has become a major problem, and the cash flow is also stretched, so they have to start the road to listing. On July 20, 2021, the AI ​​Four Little Dragons Cloud will have a successful meeting with technology, giving the industry a glimmer of hope.

For the money-burning AI track, after the meeting, listing and financing is an important match point. No corporate or quickly overcome this obstacle losses, or continue to look for external financing, while two roads which one is not easy.

01 Burn money, burn money, burn money

After multiple rounds of financing, the AI ​​Four Dragons are all eager to get more blood transfusions in the secondary market.

Established in 2014, SenseTime has completed 9 rounds of financing by 2019, with a total disclosed amount of more than 3 billion U.S. dollars, and is dubbed a financing machine by the outside world.

In the following two years, SenseTime did not have any new financing. Is there no need for external blood transfusion? not necessarily.

There is indeed some news that SenseTime has achieved profitability in 2017. A financing plan posted on the Internet shows that SenseTime’s net profit in 2017 and 2018 will be 11 million yuan and 59 million yuan. However, net profit is not equal to net cash inflow. After all, the other three companies have the characteristics of long collection of sales and high accounts receivable, and SenseTime is no exception.

Until the end of 2020, SenseTime has completed a round of Pre-IPO financing with a post-investment valuation of US$12 billion. The specific financing amount was not disclosed.

In June of this year, it was reported that Shangtang Technology opened the A+H share listing and will submit a prospectus to the Hong Kong Stock Exchange in August. The news about Shangtang’s listing has been circulated for several rounds. This time it is a vague answer and no comment.

The Megvii technology, which is also the AI ​​four dragons, is not much better.

Megvii’s last round of financing stayed in May 2019. From 2013 to 2019, Megvii’s accumulated financing amount was nearly 9 billion yuan.

In less than four years according to the prospectus, Megvii’s cumulative loss was 13.06 billion yuan. Excluding the influence of preferred stocks, the total loss was about 2.9 billion yuan, plus some cash expenditures for investment activities, 9 billion yuan. flower.

With the slow progress on the market, according to data in its third quarter of 2020, monetary funds and capacity on account of volatile now trading financial assets add up less than 28 billion yuan, according to the rhythm of business investment in 2019 and the progress of business activities and The net outflow of investment activities is about 2.5 billion yuan. If there is no new capital entering, according to the previous burning method, the 2.8 billion yuan can be maintained for one year.

Therefore, in addition to the 4.75 billion yuan used in construction projects, more than 1.2 billion yuan will be used to supplement liquidity in the proposed listing of Megvii Technology.

Look at Yitu Technology, which has terminated its listing. Last year, there were news of salary cuts, layoffs, and year-end bonuses. The plan was to supplement liquidity with 2.238 billion yuan of funds raised after the listing. 2.238 billion yuan, which is nearly 700 million yuan more than the funds on Yitu Technology’s books on June 30, 2020. The pressure on funds is evident.

Compared with the previous companies, Yuncong Technology’s funding pressure is much smaller. Before listing, there were 3.5 billion yuan of financing, and the total loss in the past three years was also relatively low, about 1.3 billion yuan. Fortunately, I got the admission tickets for the secondary market. It is expected to raise 3.75 billion yuan, of which 700 million yuan. Used to supplement working capital.

Capital is not charity, it is profit-seeking. Promoting the listing of companies is a way to withdraw from capital, and it is also a means to realize capital appreciation. Some companies will sign a gambling agreement with a company to be listed. If they cannot be listed, they have to fulfill the obligation of share repurchase, which can be regarded as a kind of real stock debt. Of course, these will not be exposed.

This will undoubtedly make the cash flow of some companies even tighter.

02 Why do we need to burn money?

There is a saying in the industry that 9 of the 10 AI companies are losing money, and one is filing for bankruptcy. One sentence accurately describes the burning characteristics of AI companies.

So where did the money go?

The essence of the business of the AI ​​Four Dragons is to maintain the leading edge of technology, and then explore the scene-based implementation of technology.

Let’s look at the first point. To maintain the leading edge of technology, we must continue to invest in research and development. Cambrian’s R&D investment is higher than operating income. Yitu Technology’s R&D accounts for more than 90% of its revenue, and the proportion of Megvii Technology and Yuncong Technology is also more than 70%.

The Deloitte report shows that the artificial intelligence ecosystem is divided into three layers: infrastructure, technology platform, and landing scenarios. Infrastructure mainly includes cloud computing platform providers, chips, IT software and hardware system providers; technology platforms mainly include artificial intelligence algorithm companies, machine learning platform companies, knowledge graph technology providers, intelligent voice companies, RPA providers, and AI software frameworks ; Landing scenes include a large number of rich scenes such as industry, finance, medical care, and retail.

Let’s look at the second point, the scene is implemented.

In the early stage of AI development, technology was used to find landing scenes, that is, holding a hammer to find nails. Now the way of playing has become the reverse to build a suitable hammer according to the shape of the nail.

Every AI company has different industries downstream. Take Yuncong Technology as an example. Finance, security, transportation, etc. all have customers to serve, and there are more application scenarios.

The AI ​​models and underlying algorithms for different scenarios are different. How to make the technology and operating system best fit a certain scenario requires continuous exploration.

The current artificial intelligence is in the early stage of development, and the iteration speed of customized solutions for related technologies and application scenarios is also relatively fast. Taking Yuncong as an example, its product iteration cycle is generally only 2-6 months. Therefore, R&D in the AI ​​industry is a long-lasting and high-investment process.

This is also the reason why the technology is difficult to implement. In addition, there are obvious differences between projects. There are more customized products than standardized products, and the cost will inevitably go up. It seems that AI is an industry with high gross profit margins, but in fact it is blood loss.

Finally, the technology landing of the exploration scenario ultimately depends on marketing.

Technical superiority is the cornerstone, but if the technology is severely disconnected from the market, it will be a fatal blow to commercialization.

Having a correct and clear understanding of the needs of potential customers, being able to carry out creative design and promotion, and making timely adjustments to products with poor sales are all important aspects of product marketization, and sales expenses have also become important expenditures.

In 2019, the sales expense ratios of Yitu Technology, Megvii Technology, and Yuncong Technology were 58.31%, 27.04% and 28.29%, respectively.

Therefore, no matter which link, you have to invest heavily.

03 Is it possible to make a loss?

When is the head of burning money?

Speaking from the source of the AI ​​four little dragons, they all started from the field of computer vision, because of the homogenization of application scenarios, such as personal combat in the security field, and encountering big players in the traditional security field, the competition has become particularly fierce, and the blue ocean Shengsheng was beaten into the Red Sea, after which the AI ​​four dragons began to move into different subdivisions.

Yitu Technology has chosen a heavier path, AI chip + computing power manufacturer, using algorithms to redesign the chip architecture, has its own advantages in medical imaging; finance and security are still the focus of Yuncong, and now it has extended its wisdom Business and smart governance; Megvii Technology cuts into the AIoT field and regards the Internet of Things business as a future growth point; SenseTime has laid out its business lines, from the AR function of beauty to AI education, self-positioning as an “AI factory “.

Now they stand in different vertical fields. The same thing is that they all hope to reduce the marginal cost by building standardized and reusable products and services, and the integration of software and hardware.

In other words, we must make standardized products in differentiated industries.

Then, the key to the problem is whether there are enough differentiated industries. Are there late entrants and will the fight in the security field be reproduced?

Megvii Technology mentioned in its prospectus for Hong Kong stocks: The industry we operate in is very competitive and faces competition in multiple business areas.

The current competition includes four levels: competing with other companies that focus on the development and commercial production of artificial intelligence technology; competing with existing participants in the vertical fields that are not focused on artificial intelligence; competing with new industry entrants ; Potential competition with global technology companies.

The prospectus of Yitu and Yuncong also mentioned that they have to compete with traditional security companies, IT giants and Internet companies.

The “Research Report on China’s Growth AI Companies” pointed out that there are many giants in the field of artificial intelligence, and leading companies in the industry tend to open market frameworks or algorithm technology platforms to provide basic support for more companies, while some market scales are not large. Refining the scene, the giant companies will basically not get involved.

The industry is not big enough, the development of the enterprise itself is restricted, and the industry is big enough, and it is easy to attract giants. The former situation makes it difficult to achieve standardization to reduce costs, and the latter situation may turn the industry into a red sea again before hematopoiesis is achieved.

Posted by:CoinYuppie,Reprinted with attribution to:https://coinyuppie.com/ai-four-little-dragons-the-death-of-unbearable-cash-flow/
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.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-07-22 08:40
Next 2021-07-22 08:42

Related articles