The smartphone industry is ushering in a new inflection point.
According to data released by IDC, the shipment volume of China's smartphone market in 2023 will be about 27.1 billion units, down 50%, the lowest shipment volume in nearly 10 years. However, there are also signs of improvement, with the fourth quarter showing its first growth in 10 quarters.
The new recovery cycle also means new competition. In the past few years, mobile phone manufacturers have "rolled" up in hardware, such as new chips, high brush screens, images, fast charging, etc., which are now the standard configuration of flagship models. In order to further improve the user experience, hardware stacking can no longer meet the needs of users, and the development and innovation at the software level can adapt to the new situation.
Starting from the end of 2023, the concept of AI device-side large models is being installed on various models, and its intelligent capabilities have brought more possibilities to the smartphone industry.
On the one hand, the gradual evolution of AI is expected to revolutionize the experience of using smartphones and drive more consumer demand. On the other hand, AI capabilities will become a powerful force for mobile phone manufacturers to compete in the market, and it is an important bargaining chip to gain an advantage in the new battle.
It's a battle for the future around user experience, and a pivotal battle for the next market landscape.
Beach-grabbing AI model
Bill Gates once asserted that "ChatGPT is as important as the invention of the internet and will change the world." ”
The essence of this change is to change the way humans and machines interact, and now it is gradually becoming a reality in the way that AI empowers smartphones.
On January 18th, Samsung's new generation of flagship S24 series was officially released, and AI capabilities were taken as the focus of this build, shouting the slogan of "opening a new era of mobile AI", and the new machine introduced a number of AI functions such as **AI processing, AI chatbot, image processing, and real-time call translation.
There is also news that Apple will introduce AI in iOS 18, which will be released next year, giving Siri more intelligent and natural conversational capabilities.
In fact, in the layout process of this AI model, domestic mobile phones have taken a step ahead.
In the recent flagship launch wave, vivo's blue heart model, OPPO's AndesGPT end-side model, and Honor's magic model have appeared one after another, bringing a variety of innovative features.
Behind this, on the one hand, there is a surplus of hardware performance and a slowdown in innovation, and on the other hand, let the competition return to the essence of the user experience, mobile phone manufacturers pay more attention to the creation of soft power than before.
Observing the development context, the advancement of AI capabilities on mobile phones is a gradual deepening process, which can be roughly divided into three stages, namely:Initiation, growth and deepening.
In the initial stage, AI can only implement some simple scenarios, such as the well-known voice assistant, which can cope with simple Q&A and control; It can support the recognition of different objects on the image.
In the growth stage, AI can initially learn the user's habits of using mobile phones, and give recommendations that are more suitable for the scene based on location, time, and previous habits.
At present, the development is accelerating, and people have more complex requirements for AI capabilities. Vendors are trying to take the lead in introducing AI large models into the device side to achieve multi-modal natural interaction, accurate intent recognition and other experiences, and the information and interaction forms output by AI to users have become richer.
The user's direct experience of this is that it "understands me" better, and AI can conduct more complex conversations, instructions, create more expected pictures, and solve specific problems in life to improve life efficiency.
For example, in terms of voice, text, image and ** processing, AI can translate in real time during calls; According to natural language interaction, complete the search**, and complete the operation, such as moving, eliminating or generating similar images for filling; Or take it a step further and quickly clip and generate personal **.
However, it is not new for mobile phones to be equipped with AI capabilities, and third-party AI platforms at home and abroad, such as ChatGPT, Wenxin Yiyan, Tongyi Qianwen, Tiangong and other general large models, can be experienced by most APP users.
AI-enabled mobile phones are still in the stage of groping across the river, but the changes that may be brought about in the future are enough to look forward to.
The race for user experience
New models are available every year, and whether there is an innovative innovation in the user experience is the core factor affecting the user's willingness to change the machine.
According to TechInsights, the global smartphone replacement cycle in 2023 will be extended to a record high of 51 months, which means that many users are using models from four years ago and are reluctant to replace them.
The reason is that in recent years, in addition to the folding screen that needs to be popularized, the industry's innovation is concentrated in the cliché fields such as performance, photography, fast charging, and battery life, which is the icing on the cake for the improvement of the user experience, and cannot really stimulate the needs of consumers.
There has been a point of view circulating in the mobile phone industry:
Hardware determines the lower limit of the user experience, and software determines the upper limit of the user experience.
The craze of AI large models is essentially the competition between mobile phone manufacturers around user experience.
The worthwhile part is:And why are most vendors more committed to using device-side rather than cloud-based models? And what is the difference between using a large model on the computer and using a third-party AI intelligent assistant APP?
First of allCost considerations。The vice president of vivo has said that the average cost of calling a cloud large model is currently 12 points - 15 cents, roughly calculated by the number of more than 100 million users, the annual cost may reach tens or even tens of billions of yuan. Obviously, this is not a long-term solution.
In addition, the point is:Intent comprehensionwithPrivacy and securityTwo ways.
In order for AI to better serve users, it needs a process of learning and data processing to fully understand user habits and intentions. Compared with third-party platform applications that are limited by self-starting, the system-level device-side AI model can better understand users' daily habits and gain insight into immediate needs.
Although the cloud model has an absolute advantage in computing power, there is a large amount of private data in the user's mobile phone, such as personal information, financial information, etc., and there will be security risks if it is uploaded to the cloud.
Putting the AI computing process on the device side can protect privacy from the root and provide a secure foundation for AI to evolve into a smart assistant, so that people can buy and use it with confidence.
However, it is still too early to see AI as an innovation that will disrupt the smartphone experience.
AI computing occupies a huge amount of processing resources, and the computing power of mobile terminals needs to be improved, and adaptation and optimization are also long-term tasks. In addition, in the actual experience, the current AI intelligence is not as smart as expected, and the application effect of various services is often discounted compared with the publicity.
For example, the value of the answers provided by AI is limited, and the efficiency of obtaining valid information is not very high; The recommendation service function provides a convenient entrance based on comprehensive factors such as positioning, time, and habits, but it does not fully understand the user's intention; Not only is there always a flaw in the creative function, but it also makes jokes from time to time. And so on and so forth.
Time will tell how far AI can make a difference. What is determined now is that AI empowerment has enabled the mobile phone industry to see a new direction of innovation, which has become a major trend in the future. According to Counterpoint Research's Generative AI Smartphone Shipment Insights** report, 2024 is expected to be a pivotal year for generative AI smartphones, with shipments expected to reach 100 million units.
What is certain is that for mobile phone manufacturers, large model mobile phones are not only a new selling point to create differentiated advantages, but also a powerful way to highlight the high-end image of mobile phone brands.
The battle for the future
Competing for AI is the trend of future competition, and to use AI as a powerful command, it requires in-depth cooperation at the system level, and at present, manufacturers are creating a combination of "AI + self-developed OS".
In the past 2023, Chinese mobile phone manufacturers have either developed their own or re-developed the underlying based on the Android kernel, and have launched more independent mobile phone OS.
In October last year, Xiaomi MIUI, which has many fans, finally transformed into a surging OS, and was the first to be carried on the Xiaomi 14 series; Although Vivo's subsequent release of Blue River BlueOS is only for peripheral ecological devices such as smart watches, it is decoupled from Android and can accumulate valuable experience; With the help of the "ColorOS supercomputing platform", OPPO has rewritten the underlying framework of Android.
Honor's iterative upgrade of MagicOS has achieved stronger independence and richer ease of use; On January 18 this year, Huawei announced that HarmonyOS Next (HarmonyOS Next version) is no longer compatible with Android applications, and after 4 years of development, it has finally won the complete body, which is also a leapfrog step for domestic OS.
Behind this, domestic manufacturers are aiming at autonomy and differentiation.
With Huawei's experience of being controlled by others, manufacturers have realized that the foundation is in the hands of others, and it is inevitable to look at people's faces. Since then, Huawei Hongmeng has taken a path of independence, which is also a great encouragement for domestic brands.
In addition, deepening the cultivation of independent OS is also the only way to create a differentiated experience. Brands have tried to create novel features, but they are often limited by the limitations of the Android framework, making homogeneous features, which has attracted questions about the lack of innovation. Only by building the integration of software and hardware, and at the same time enhancing the ability of ecological integration, can we create differentiated selling points as much as possible, keep our basic market, and face more fierce competition.
In the long run, under the trend of the Internet of Everything, mobile phones are also supercomputing centers for cross-platform operation.
In the era of the Internet of Everything, the complexity and number of operating system branches involved in various devices, and the ecological connection gap between different systems and protocols may exceed many people's imagination. As Lei Jun, the founder of Xiaomi, mentioned at the release of The Paper, in order to make a wide layout in the hardware ecology of mobile phones, PCs, smart homes, and even cars, it is necessary to have an operating system that can truly realize multi-device collaboration.
The in-depth combination of AI and self-developed OS can connect various in-depth scenarios and launch corresponding innovative products, which is also the only way to break free from the shackles of hardware stacking and hit the high-end.
IDC data shows that in recent years, the domestic smartphone market has continued to rise in the high-end market share of the price segment above US$600 (corresponding to 4,000 yuan), while the terminal market has shrunk, which means that high-end models are increasingly recognized, and users are willing to spend money to buy a value-for-money experience.
Behind this, the performance of domestic brands is becoming more and more eye-catching. In the past, innovation was more concentrated on Apple and Samsung, whether it was the appearance design, hardware configuration or system experience, they were highly regarded objects, and "iPhone moment" has become a common term for all walks of life to advertise their own products.
However, in recent years, the two major manufacturers seem to have fallen into the trap of "squeezing toothpaste", and domestic manufacturers have accelerated the pace of catching up. Under the guidance of technology iteration and market demand, the new round of competition pattern may have a different development.
Conclusion
With the update and iteration of AI technology, the functional experience on mobile phones is facing innovation, intelligent functions and personalized services have become the focus of new competition, and the entire industry is also ushering in new opportunities.
Only by constantly innovating, meeting the needs of consumers, and grasping the development trend of new technologies, can we find new growth points in the cold winter of the industry and realize the leap in value.