GuShiio.com Wen| Cultural Value Officer, Author| Hoshino, Editor| Meiqi
The battle for AI applications is rapidly moving from a battle for products and technologies to a battle for platform ecology.
In recent times, Yuanbao’s advertisements have been swept across almost all platforms, including Tencent App. Even WeChat, the coldest WeChat in Tencent’s department, has become a diversion tool for Yuanbao, and has even entered the Jiugong Grid of life services in one step. It can be said that Tencent is gathering the efforts of the whole family to help Yuan Bao reach the top in the AI battle.
In the face of the fierce Yuan Bao, Doubao, who has been in a passive state this year, has finally begun to fight back. According to Tech Planet, Douyin App recently tested the AI capabilities of accessing Douyin App, and opened two super entrances for Douyin App within the Douyin App.
It can be seen that Byte is giving up the strategy of making bean buns a super application on its own, and the ecological integration between AI applications and super platforms has become a new competition point in the next stage.
After upgrading the brand Slogan to an AI all-round assistant for 200 million people in January, Quark officially became a clear card for Ali AI to C. At the end of February, Quark AI search went online and thought deeply, officially standing on the same starting line as Doubao, Yuanbao, etc. In the next step, the deep integration of Quark and Ali products and content ecosystems is also worth looking forward to.
The giant gave full support,Which AI application can become a super portal?
If we compare the three AI applications of Doubao, Tencent Yuanbao, and Quark horizontally, who is most likely to form their own ecological moat and become a strong contender for the super entrance of the AI era?
Since its birth, Doubao has inherited the entertainment gene of bytes, encouraging institutions and users to create an agent ecosystem through PGC and UGC, covering entertainment, games, learning and other scenarios. However, the successive attacks of Deepseek and Tencent Yuanbao proved that the intelligent ecology of Doubao did not constitute a moat.
Most agents are just anthropomorphic and vertical packaging of AI conversations, meeting users ‘lightweight entertainment needs, and there is no irreplaceable user stickiness. Moreover, intelligent agents that are not smart enough will only cause users to go from disappointment to loss. One developer revealed that there may be only a few hundred active agents with more than 400,000 popular agents. Strong streaming has led to the short-term influx of users being too general, but Doubao has no just-needed application scenarios to leave it behind.
Compared with bean buns, Tencent Yuanbao has deliberately strengthened the tool attributes and has now integrated lightweight tools such as AI reading, AI retouching, and translation. On March 1, Tencent Yuanbao launched the PC version, which means that it has realized that the computer side is inaccessible for it to become a powerful efficiency tool.
However, the user retention capabilities of these low-frequency tools are still in doubt. Tencent Yuanbao’s core competitiveness still remains in the simple superposition of Tencent’s traffic +DeepSeek technology. This model can easily be replicated by competitors.
Compared with the AI dialogue AI agent path of bean buns, Tencent Yuanbao’s AI dialogue lightweight application path, Quark’s evolution path is just the opposite. Accumulates product and service ecosystems from one application scenario to naturally add AI capabilities in the era of big models. Improve the ecology and successfully transform users of intelligent search, online disk, search questions, scanning king, etc. into AI users. These are high-frequency requirements that have been verified by hundreds of millions of users, rather than pseudo-requirements imagined from the perspective of AI technology. Their cohesion to the AI ecosystem and users obviously cannot be underestimated.
Before the All in AI strategy was proposed, Quark was not regarded as a member of the AI application track. As a low-key player with 200 million users, before this wave of AI arrived, Quark’s potential was not fully recognized, both inside and outside Ali. Nowadays, Quark, who has transformed from AI search to AI assistant, may be one of the most seeded players that cannot be ignored.
Super AI product factoryOr is it an open AI application ecosystem?
Of course, Byte is not unaware of the importance of high-frequency AI tools in improving user Retention rate. According to media reports, there are multiple AI product business lines and services within Byte that are being pushed forward, including Rubik’s Cube, Jiyi, Xiaohe AI Doctor, Youma, Flinco, etc., covering C-end and B-end, involving information reading, medical care, programming, AI basic technology and other fields, a super AI product factory has been opened.
Obviously, Byte is no longer satisfied with the entertainment attributes of bean buns, but hopes to create the next App Store in the AI era. However, compared to the agent ecosystem created by the efforts of users and product managers, the imagination space of product factories inside Byte is relatively limited.
While going to the field in person, how to stimulate the enthusiasm of external developers and create an open and win-win AI application ecosystem is an important topic that Byte has to overcome.
In contrast, Tencent Yuanbao’s AI application has introduced service providers such as Tongcheng Travel and Express 100. If it is deeply integrated with WeChat in the future, all service numbers can be empowered by AI and evolved into smarter services.
Of course, to achieve this, Tencent will need to take drastic steps to tear down the high walls between departments, break the relatively independent state of WeChat, and even require in-depth adjustments in the organizational structure.
Left:In bean buns:Tencent Yuanbao right:quark
Starting from a browser and quickly seizing users ‘quarks with vertical application scenarios, in addition to being able to target development for scenarios such as college entrance examinations, it also has a stronger open gene. From the perspective of the underlying logic of the product, Quark is obviously different from AI products such as bean bags and Tencent Yuanbao. It does not concentrate all functional entrances into a single search/dialogue entrance. It retains a large number of tool entrances such as documents, online disks, and scanning. Users have long become accustomed to using it as a universal toolbox.
In addition to a small number of tools developed by the official team, most of Quark’s current toolbox call third-party products and services, which has formed an extremely rich developer ecosystem. In the next step, if AI capabilities are opened up to all service providers, so that the search portal is not just an information portal, but also an portal to invoke services across the network, it will be a matter of course.
Once upon a time, from light applications to Mini programs, Baidu tried to directly complete high-frequency services such as hotel booking, train tickets, and movie tickets within the App to realize the interconnection of network services, but it failed. The new wave of AI may make such ideals a reality, and Quark is one of the most promising competitors.
The reason why I am not optimistic about Baidu is because Baidu’s user impression as a search engine is deeply rooted. From returning to Baidu, you can also see that the classic Sloga, which you know, Baidu’s positioning is still an information and entertainment platform, rather than a product and service platform. The two failures of light applications and Mini programs may have made it no longer dare to try it easily.
“After Dali’s miracle failed,C-end commercialization problems to be solved
At present, the battle for AI applications is still at the stage of vigorously creating miracles and racing around the land regardless of cost. Some media learned from the AppGrowing platform that as of February 27, Tencent Yuanbao spent a total of 281 million yuan in 27 days on launch.
However, the eclipse of Kimi, the king of investment last year, makes us have to question mark this kind of money-burning competition. Once this round of AI craze gradually recedes and users from early adopters are lost, how can AI applications that have not yet formed blood-making capabilities continue to advance, and how long can the giant’s infinite bullets support it?
Putting aside the increasingly cruel price war in the B-end market, the current profit prospects of the C-end AI application are inseparable from high-speed user growth. Only a huge user base can lay the foundation for subsequent value-added services and accurately matched advertising.
However, when the moat is unstable, no company dares to take the lead in commercialization. In a sense, this is similar to a classic prisoner’s dilemma.
In fact, commercialization, especially the user-paid model, may be the best way to test user needs, and it can also prevent the platform from being blinded by the false prosperity of burning money to buy volume.
Bean bags are mainly realized through paid plug-ins such as hot copy generation on the C-end. Douyin creators generate video scripts through bean bags, which may have better traffic performance on the platform. However, this monetization model that relies on one’s own platform currently seems to be a little single.
Tencent Yuanbao, which is in a period of rapid expansion, should not bear the pressure of commercialization in the short term, but once users penetrate into the bottleneck period, they may eventually face torture on their profit model.
At present, Quark has already realized part of its monetization through member fees and other channels. Moreover, as a content platform where users can browse news and follow up novels, Quark’s advertising commercialization potential needs to be further released. The reason why the membership payment model can be implemented also proves that the value of this collective toolbox has been recognized by users. After enhancing AI capabilities, users ‘paying propensity is expected to increase further.
Whether it is bean buns, Tencent Yuanbao or Quark, they are all springboards for mobile Internet giants to seize the next AI era. How far they can help giants jump? In addition to the strong support of the platform ecosystem, it may also depend on the elasticity (possibility) of the springboard itself. How big is it? As the development paths between the three companies become increasingly clear and differentiated, we may soon find out.
