Before making the Doubao AI phone, Zhang Yiming should also have a chat with Ma Huateng.

The Doubao AI phone has become a hit. The original price of the Doubao phone is 2,499 yuan, but on Xianyu, it can be resold at a price as high as 8,000 yuan. This price increase rate is comparable to that of the iPhone 4 when it was first released.

The popularity of the Doubao AI phone lies in its high - privilege Agent, which enables system - level AI capabilities. Users only need to say a single sentence, and the AI can easily perform various complex operations, such as comparing prices and placing orders on multiple e - commerce platforms, calculating red envelopes to order takeaways. These are just basic operations.

As a result, people began to exclaim: All the big Internet companies and super apps are going to be eliminated.

Some time ago, Elon Musk said, "Phones and apps will be obsolete in five years." If Musk's prediction comes true, then the AI phone might just be the beginning.

Soon, there was a new development.

Some users found that they couldn't use the Doubao AI assistant to operate WeChat. Subsequently, some media reported that the "Software License and Service Agreement" of WeChat clearly prohibits automated operations by third - party tools.

Tencent has long been on guard against AI.

Although whether users can send messages on WeChat through the Doubao AI assistant in the future may depend on Tencent's attitude, the "sudden emergence" of the Doubao AI phone has at least two major significances:

First, it once again verifies that the path of AI interaction in the mobile phone industry is correct.

The popularity of the Doubao AI phone verifies the strong certainty of the investment of major mobile phone manufacturers such as Honor, vivo, and OPPO in the AI direction: AI interaction partially replacing screen interaction is the major trend in the future.

Second, the popularity of the Doubao AI assistant this time is an important "trigger point", and its significance is no less than that of Mark Zuckerberg's personal attendance at the Oculus press conference in 2016.

According to the financing information on Tianyancha APP, at that time, Mark Zuckerberg had acquired Oculus for $3 billion. After Zuckerberg personally appeared at the company's press conference, the VR market was completely ignited.

Today, the Doubao AI has also detonated the AI phone market. However, how far the Doubao AI phone can ultimately go? It may take time to give an answer.

The "Revolutionaries" and "Conservatives" in the Implementation of AI Phones

Can the Doubao assistant overthrow traditional mobile phones?

Currently, it seems not.

There are several reasons:

1. Mobile phones are products that integrate software and hardware. ByteDance is strong in software technology, not hardware capabilities. Even ZTE, the partner of Doubao AI, has been left far behind by players such as Huawei, Xiaomi, OPPO, vivo, and Honor in the mobile phone industry.

Realistically speaking, even with the super - strength of Doubao AI, ZTE still has to face the reality of being "unable to compete" in the mobile phone industry in the short term.

2. Developing system - level AI is a gradual process, and ByteDance may be moving too fast.

ByteDance is not the only one working on operating system - level AI. Many manufacturers are doing it. Currently, there are roughly two camps in the market:

The radical revolutionaries and the moderate conservatives.

A typical representative among the revolutionaries is the Doubao assistant. Why has the Doubao AI phone sparked so much discussion? It's because system - level AI fundamentally affects user habits and challenges the underlying logic of the traditional mobile ecosystem.

What is the most fundamental logic of the Internet mobile ecosystem? It is to occupy the traffic entrance and then monetize through advertising and e - commerce.

AI's cross - app operations directly affect the user time of various e - commerce apps and group - buying apps. Users can directly and accurately complete operations without watching ads or browsing the home pages. That is to say, once system - level AI emerges, theoretically, the home - page advertising recommendation mechanism of big - company apps may become ineffective.

In the eyes of big companies, this is not a "sharing the cake" logic, but a blatant "turning over the table": There will be no more e - commerce entrances or takeaway entrances. In the future, users will only have one AI entrance.

Why did WeChat ban AI automated operations so quickly?

On the one hand, it is because the user agreement already prohibits it. On the other hand, from a business logic perspective, if AI can directly help users send messages on WeChat, WeChat's DAU and MAU may also be affected.

If the DAU and MAU data are not good, how can it sell ads?

It's not just WeChat. In the face of AI, all tool - type apps will face such an awkward situation. The online business logic of payment, e - commerce, and local life may have to be rewritten.

Therefore, the implementation of system - level AI is not only a technical issue but also a distribution issue. This is also the most difficult part for the revolutionaries. The more intense the "revolution" is and the more interests it affects, the stronger the opposition it will face, and in the end, the radical revolution may not succeed.

Compared with the radical revolutionaries, the relatively moderate "conservatives" may be more likely to get the job done.

First of all, mobile phone manufacturers are in a relatively neutral position. Smartphones are terminals and the foundation of all software. From an implementation perspective, to achieve this goal, it must be a product of mutual compromise and support between software and hardware, and a compromise route needs to be taken.

Mobile phone manufacturers have their own application ecosystems, have long - term cooperative relationships with major app manufacturers, and have no direct business competition. It may be easier for other big companies to accept if mobile phone manufacturers promote it.

Secondly, from a technical perspective, today's Doubao AI phone has obtained system - level code permissions. However, the real AI transformation of mobile phones is not just about opening system - level entrance permissions, but about a truly AI - native underlying operating system.

The ones capable of doing this are not large - model manufacturers, but mobile phone manufacturers with system R & D capabilities. There are actually only a few mobile phone manufacturers in the market with system - level development capabilities: Huawei, Xiaomi, OPPO, vivo, and Honor.

Finally, many mobile phone manufacturers are already working on system - level AI.

Taking Honor as an example, on the Magic 8 series, Honor has achieved AI - automated scenarios. Functions such as ordering takeaways and buying things with AI can all be realized, and the overall experience is not bad.

Moreover, in recent years, many mobile phone manufacturers have been fiercely competing in AI capabilities.

Therefore, for ByteDance, directly launching the Doubao AI phone may mean "making enemies" in the mobile phone industry, and it may not be the most advantageous strategic choice.

The Doubao AI phone is still too dreamy and idealistic, giving a feeling of being a bit unrealistic.

It's like seeing a beautiful concept car at a car show. The design is perfect and it represents the future development trend of the industry. Maybe it can't be mass - produced in the end and won't be popular in the mass market, but it points the way for the future.

ByteDance is an app factory and an aggregator of algorithm technology. However, making an AI phone is not just about creating a product, but about building a new industry ecosystem that integrates software - hardware and involves cross - field cooperation. How to build this ecosystem may be the key to solving the problem.

The "Dilemma" of AI Phones May Be Solved by Ma Huateng

Why are mobile phone manufacturers more likely to succeed in developing system - level AI for mobile phones?

I think there are several important factors.

First, experience in developing software - hardware integrated products.

The reason why only mobile phone manufacturers can integrate AI into the underlying mobile phone operating system is that developing software and hardware are two different things. There are indeed not many companies in the world that can understand both.

Historically, giants that transitioned from software to hardware have all failed.

Microsoft has very advanced technology. When Microsoft tried to enter the mobile phone market, even after buying Nokia and developing a mobile phone system, it still ended up in failure.

Another example is Google. Google has developed many cutting - edge technologies over the years and has also made great achievements in the AI field. But what is the global market share of Google phones today? Only 3%.

In this sense, ByteDance is good at software, but it may not succeed in making AI phones.

Second, whether AI can succeed depends not only on the strength of the technology but also on whether an ecosystem can be built.

After the Doubao phone became popular, many tech industry experts spoke up. Luo Yonghao said that technological revolutions are unstoppable. Fang Fei from Honor said that the development direction of Doubao is highly consistent with that of Honor, and he is looking forward to seeing more partners join.

Honor has been exploring the field of AI phones for some time. ByteDance's focus on the Doubao AI phone is, in a sense, a "cross - verification" of the AI strategies of mobile phone manufacturers.

From the perspective of mobile phone manufacturers, the hotter the AI phone track is and the more topics it generates, the more benefits mobile phone manufacturers will get.

After the launch of the Doubao AI phone, ZTE's stock price soared. In fact, not only ZTE, but the valuations and market values of companies related to the AI phone concept may also increase significantly. The reason is simple. For mobile phone manufacturers, a single manufacturer represents not only a brand but also an entire ecosystem.

Still taking Honor as an example, in March this year, Honor launched the "Alpha Strategy", which aims to build a globally leading AI terminal ecosystem in three steps. The first step is to "build smart phones". According to Honor's plan, next year, Honor will launch a Robot Phone that integrates AI phones, embodied intelligence, and high - definition photography.

In fact, the Alpha Strategy proposed by Li Jian is not only an AI strategy but also an ecosystem strategy. In the implementation of the strategy, many of Honor's actions are centered around the industry ecosystem. For example, MagicGUI is open - sourced to provide a "general AI foundation", and developers can directly develop cross - device and cross - application AI functions.

Li Jian also said, "A real AI OS needs to complete a value leap centered around people... User co - creation, industrial symbiosis, and ecosystem co - prosperity together constitute the three driving forces for the evolution of Honor Magic OS."

Not only Honor, but also mobile phone manufacturers such as vivo and OPPO are building open ecosystems around AI.

Why? It's not that mobile phone manufacturers are unaware of the value of AI, but because the entrance position of AI is too important and involves many privacy and security issues that cannot be solved independently by a single company.

The problems faced by the Doubao AI phone are similar to those in the implementation of L3 autonomous driving.

Is it because the technology for L3 autonomous driving is not available? Actually, the technology has been available for a long time. It's more of a matter of safety costs.

Due to the existence of telecom fraud and data security issues, the implementation of system - level AI phones should actually be more cautious. Mobile phone manufacturers have experience, technology, and are more capable of balancing these aspects.

Of course, aside from these, the most fundamental thing is how to coordinate with super apps and big Internet companies. For example, in terms of business logic, it should shift from "overthrowing the big - company entrances" to empowering the industry to create incremental value.

This reminds me of when WeChat launched mini - programs.

When WeChat first launched mini - programs, people exclaimed that WeChat was going to become an "operating system" and overthrow all app entrances. As we all know later, mini - programs have instead become a tool for Tencent to release internal incremental value.

The idea of the Doubao AI assistant to serve as an entrance for services and products between the system and apps is correct. But from an implementation perspective, it can actually learn from Tencent's experience in launching mini - programs.

Moreover, in today's chaotic AI industry, the Doubao AI is not irreplaceable, but WeChat is very difficult to replace. Otherwise, when WeChat "fought" with Apple over the commission issue, so many users would not have taken WeChat's side.

Today's Doubao AI still lags behind WeChat in terms of user volume, usage habits, and user stickiness. For users, the Doubao AI is far less important than WeChat.

Therefore, what WeChat couldn't achieve back then, Doubao may also not be able to achieve.

Although it didn't become an "operating system", the open WeChat achieved another kind of success. Today, the content and traffic ecosystem of Douyin is actually no worse than that of WeChat back then.

Theoretically, the Doubao AI is completely capable of releasing the incremental value within the ByteDance ecosystem, just like WeChat mini - programs. However, today, big companies don't see an open ecosystem in the Doubao AI phone. Instead, they may see the ambition of AI to overthrow super apps and become the "entrance".

Today's launch of the Doubao AI phone is somewhat similar to the launch of apps like YYB and Wandoujia back then.

These app stores controlled the app distribution channels and were the first entrances for Internet companies to distribute apps. What was the result later? Mobile phone manufacturers have successively launched their own app stores and started to develop the pre - installed app market. All kinds of mobile phone assistants and app markets eventually declined.

Will AI phones repeat history? Time will eventually give an answer. But for ByteDance, whether to go for a closed or open model is a question worthy of further consideration.

Why was WeChat able to build an open ecosystem with mini - programs?

It's not because Tencent is good at social media. After the "3Q War", Ma Huateng decided to make a change and re - established the values and strategic tone: Tencent wants to connect everything.

Tencent's claim to "connect everything" is very valuable. Today, Tencent's market value in the Hong Kong stock market has exceeded HK$5.5 trillion.

Perhaps, today, ByteDance has also reached a stage of re - defining itself.

In the past, in the eyes of outsiders, Zhang Yiming set a tone for ByteDance that was a bit "algorithm - oriented" and "technology - focused". However, today, ByteDance is no longer a startup.

Today, ByteDance has a powerful product like Doubao AI, which is sufficient to demonstrate its technical capabilities, ecosystem capabilities, and foresight in future strategies. As an entrepreneur leading the development direction of AI technology in China and even the world, Zhang Yiming is also leading the way in the development of AI technology.

In fact, the bold innovation of the Doubao AI phone represents that Chinese AI is starting to enter the "uncharted territory". In the "uncharted territory", every small step in the implementation of AI is a big step in the history of human AI.

In a sense, the value of Zhang Yiming and Doubao AI to today's Chinese AI is no less than that of Elon Musk to the American commercial space industry.

However, for ByteDance, on the new journey, how to build a more inclusive vision and promote historical changes in AI with a more open attitude? It is still a topic worthy of in - depth study.

The AI trend is