Frequent business soup, choose to survive

Author | Chu Yilong

Source | Insight New Research Society

Stocks' ban, capital escape, stock prices cut, media singing and decline ...

This year's Shang Tang is very bumpy.

According to the 2022 report data, Shangtang Technology revenue was 1.415 billion yuan from January to June this year, a decrease of 14.3%year-on-year, 66%gross profit, a year-on-year decrease of 22.5%, and a net loss of 2.561 billion yuan after adjustment.

With continuous losses, the patience of long -termism is slowly consumed, and Shang Tang is really urgent.

At the World Artificial Intelligence Conference, the chairman and CEO of Shangtang Technology and CEO Xu Li, who ended in the World Artificial Intelligence Conference not long ago, described the AI ​​landing experience of Shang Tang, and released the four innovative artificial intelligence platforms in one breath System: Sensecore Shangtang Large Device AI Cloud, SenseAuto V2X Shangtang Magnolia Lounge Cooperation, Sensemap Shangtang Multi -Smart Smart Platform, SenseCare "Smart Diagnosis and Treatment" platform.

Obviously, Shang Tang is anxious to find a new growth point outside the main security industry, but suddenly the four areas of the four areas at the same time, the front is stretched, can Shang Tang "take care of this" instead of "losing each other"? In the face of huge uncertainty, will it "attack everywhere", but in the end, what is the granular?

01 Should stop bleeding

Let's take a look at the areas of these four platforms.

Sensecore Shang Tang Dajun Device AI Cloud: It is the product form that puts the large device of Shang Tang to the product form of AI Cloud, and opens AI deployment capabilities to business users.

Simply put, under the empowerment of AI cloud, companies can realize intelligent management of AI infrastructure without the need to invest in the early stage. It aims to provide a set of developed AI infrastructure for the industry.

SenseAuto V2X Shangtang Magic Car Road Cooperation: Through the integration and analytical analysis decision -making plan of the car end and road information, the "car -road -cloud" integration decision control is achieved.

The purpose is to make smart driving vehicles more comprehensive perception ability and enhance the driving ability of complex traffic scenarios.

Sensemap Shangtang Multi -Body Smart Platform: It is a platform -type product specially dedicated to game production and operating quantities. It is mainly auxiliary game development and provides one -stop game production services.

Shangtang SenseCare's "Smart Diagnosis and Treatment" platform: It is a smart service that provides overall solutions covering multiple scenarios in the future hospital. In fact, from the perspective of product forms, it is mainly for doctors to provide effective intelligent auxiliary tools for their clinical diagnosis and treatment. Or integrate the hospital information and help the hospital managers better make decisions.

From the positioning of the above four platforms, it is not difficult to see that Shang Tang tried to establish a more efficient and easy -to -use platform service through standardization, promote the landing of AI in different scenarios, and can achieve revenue from it.

As Xu Li said: while looking up at the broad and boundless imagination of cutting -edge technology, AI must also combine with the real industrial economy.

But the current reality is that the landing of AI is not to tell the "industrial story", but to do well.

The key to doing well includes two aspects: "technical practical" and "value formation".

The reason why Shangtang's main security business is "suffering from the enemy and the enemy" is Hikvision, who entered the game before, and the pursuit of a group of AI younger brothers such as visuality, Yi Tu, Yun Cong. The technical capabilities on the track did not open the gap with friends and merchants, and stayed with the business model of "algorithm authorization", which was slightly awkward in the play.

What you can do, everyone can do it, not only "technical practical", not to mention "value formation", and the moat has never been established.

The four major platforms released by Shang Tang this time are all industries with great market potential and high return. It can be seen that Shang Tang was deeply trapped in the bottleneck of business development, eager to open up the second front line, and proves himself in the question of sang decline.

Returning to blood is already extravagant, bleeding is a more urgent reality.

The Shengjing of the Waic Double Horse dialogue was no longer. After the Capital's AI companies who told love to love stories, they also began to return to reason.

Nowadays, the commercial soup that is trapped in the predicament is to do the way to promote the ecological win -win situation in the process of promoting AI in the process of appropriate scenarios.

02 The next step of Shang Tang

AI native cloud, is Shangtang played?

Shangtang released AI native clouds, and it was more announced to the outside world that they had this ability.

Make AI natively in the cloud is the common force of the current cloud computing manufacturers, and it is also a further fusion trend of cloud computing and AI.

The contest between cloud computing manufacturers is how to meet the needs of developers' "R & D efficiency" and "delivery efficiency".

The Shangtang AI cloud can be reduced inward, but the competitiveness that meets the needs of developers is weak.

The developer is the core force of the development of cloud computing ecology. Only by the developer continuously converge on the cloud computing platform to develop various types of applications, can cloud computing ecological cakes continue to make bigger opportunities.

As far as Yu Shangtang is concerned, ecological construction ability is a problem. The bigger problem is that AI native clouds are far water, and they cannot solve their thirst.

As we all know, the cloud computing business has a large investment in funds and a long cycle. It is a business that must reach a certain scale to form a positive circulation. Can the volume and resources of Shang Tang afford it?

Shang Tang is obviously impossible. Cloud computing not only needs to understand user needs, but more importantly to create corresponding products.

The demand between different industries is very different, and the Shangtang AI cloud cannot be used to use the existing AI models for different enterprises. The ultimate situation is likely to be grasped together. Own. After 12 years of Alibaba, Alibaba Cloud has been suffering from 12 years. Baidu has invested 100 billion years in ten years, and its profit is still far away. How can Shang Tang be sure to catch up and surpass Baht?

Make money by coordinating the road, not only the last mile

Shang Tang's autonomous driving business is not successful.

In the whole year of 2021, the revenue of Shangtang Smart Automobile's business was only 70 million, accounting for only 4.6%of the total revenue. 8.5%.

The embarrassment of Shang Tang is that due to the particularity of the autonomous driving industry, Shang Tang has products, but it is difficult to land. Compared with professional players, it lacks project experience and data accumulation.

This year's transfer direction is also expected to attack the road.

Compared with autonomous driving, the commercial landing of car roads is much less affected by interference factors, and the market is also growing rapidly.

It is just that Shang Tang entered the game now and slowed down in the rhythm.

A car road collaboration project requires a large number of ecological forces, and all parties can cooperate with each other to complete. There are two roads in front of Shang Tang. One is to become an organizer of the ecology, led by a group of younger brothers to get the project; the other is to join, and become an ecological enterprise of a giant. Essence

However, the reality is that the two roads of Shangtang are difficult to get through.

First of all, compared with technology giants such as Baidu and Huawei, or professional players such as Qianfang and Hisense, Shang Tang, as a newcomer in the industry, lacks the experience and ability of organizing ecology. Miracle.

Secondly, judging from the ability of the coordination platform of SenseAuto V2X Shangtang, Shangtang wants to do a complete solution for car road collaborative projects. Its positioning is "leading the big brother". I can eat the cake I painted.

AI game platform is like a bubble

For Shang Tang, the AI ​​game platform can only be seen from a distance and cannot be played.

Shangtang's SenseMap is still the business of To B. The main function is to assist game development and use AI to improve the efficiency of game development.

For game developers, the purpose of developing a game AI and using the game AI is to assist the production of the game, so that the team can get convenience and convenience during the game production process.

However, the AI ​​game platform generally exists with large samples, high costs, and uncontrollable problems. For Shang Tang, which is urgent to stabilize the military heart, it is undoubtedly the cream of sugar, which will only increase the loss of Shang Tang.

Moreover, the existing game company can basically meet the needs of self -sufficiency. As a result, there is no need for large factories. The embarrassing situation that small factories cannot afford to buy.

Like Tencent, the AI ​​has been fully applied in the design of the entire game link of "Glory of the King". Netease has the "Fuxi Lab" that focuses on game AI research. At present The room has close cooperation.

Tencent and Netease Games occupy half of the industry in the entire industry, leaving only the "leftovers leftovers" that are difficult for Shangtang to make interest.

In the AI ​​era, algorithms and computing power can be used to solve the efficiency of game development, which is also the value of the current AI game platform.

But the AI ​​game platform is more like a seductive trap for Shangtang now. The AI ​​game platform obviously does not meet the stability of Shang Tang and it is difficult to achieve breakthroughs.

AI medical care is difficult to add difficulty

Finally, let's look at the AI ​​medical track. The same is the long -slope and snowy track, but the scene of AI landing is still very limited.

The case disclosed by Shangtang's official website shows that Shangtang has reached cooperation with 5 hospitals, and its business is a medical image AI.

At present, the medical image AI is the super red sea of ​​"red hair purple". It is not uncommon for a hospital to be invited by three or four video AI companies at the same time. Essence

A case worth thinking about.

IBM's Watson Health department, after six years of losses, was eventually sold, and IBM also suspended AI medical business.

It is not difficult to match medical AI technology and clinical needs. It is difficult to find the marginal effects and the business model of paid. In this regard, not only Shang Tang, but also other players have no grasp.

It is worth mentioning that Shangtang seems to have lack of confidence in the two businesses of games and medical care. Among the four major platforms released, many search engines retrieve Shang Tang's SenseMap and SenseCare "smart diagnosis and treatment". No more information was disclosed on the official website of Shangtang.

03 Standing Shang Tang at the crossroads

AI's actual application value does not need to be repeatedly described, and there has always been policy support behind it.

Under the double increase, what should Shang Tang go?

First of all, Shang Tang must consider the just -profit demand now.

The problem of technology profit is actually nothing to do with technology. Although Shang Tang has the strength and confidence of technology, technology profit is a business and economic issue.

The economic problem is that the marginal investment effect is decreasing, and the commercial soup that attaches too much attention to technology. As the technological investment results increase, more investment for new results is required, and the result is caught in the vortex of money burning. Shang Tang, who was standing on the tower, only saw the scenery in the distance, but did not think about how to reach the destination. For the industrial application of AI technology, the current technology has already entered the fast lane. Essence

Among the current AI companies, the general cost is high, the price is low, and it is also facing the peers. Especially in AI companies, there are basically no exclusive profit projects. All AI scenarios are similar to AI companies. It is a problem.

In fact, AI companies are more like Didi today. The results of Didi unmanned driving are much different from existing products in the market, and they can not earn money to cover the cost. The price increases users. The core issue of the soup is still degraded.

In addition, independent AI companies such as Shangtang have always faced the challenges of fragmented application scenarios. In different projects, the needs of customers are different, and the algorithm will be updated and the product reuse is low.

The independent AI company is facing the challenge of fragmentation of application scenarios, which derives three types.

The first is to deepen the application depth of AI technology. Through the combination of software algorithms and hardware terminal modularization, it has developed products that are more in line with market position, such as HKUST Xunfei's use of intelligent voice technology.

The second is to combine AI technology with other product platforms. For example, some China Science and Technology combine image AI and operating system platform technology.

The third is to choose to expand downstream coverage, use the technical advantages of high -efficiency research and development, and quickly develop the AI ​​model for different hardware equipment and scene requirements. This is the strategy of Shang Tang AI to empower hundreds of industry.

Shang Tang chose the most difficult road. Its efficient research and development technical advantages will inevitably bleed a lot, the wounds are repeated, and long -term pain. In addition, for different equipment and scene customization AI models, its cost is high, the input and output ratio are seriously imbalanced, and it is even more difficult. Therefore Until the market and other reasons, the four major sectors of evolution are actually walking between the hundred industry.

One of the big problems in Shangtang is confusion. Is the object of AI landing to B or to C? How to choose many scenes of AI landing, and which of the track that you mainly attacked?

At present, Shang Tang is in the stage of To B. Only by finding a commercial positioning and learning to develop the market, can you know where to go in the crossroads?

- END -