In 2026, AI hardware has found a methodology for creating blockbuster products.
In the early spring of 2026, when we re - examine the development trajectory of China's AI industry, we will find an interesting phenomenon:
After the initial parameter competition and computing power competition in the generative AI wave that started in 2023, everyone's attention is accelerating towards vertical scenarios.
On the software side, Agents represented by Manus and OpenClaw are redefining the value delivery method of AI;
On the hardware side, there are emerging small giants like Unitree and Zhipu in the robotics field, and a dark horse like Fuzai has emerged in the children's toy field. Even in the highly competitive smartwatch market, Xiaoshui Intelligence has emerged, with an annual shipment of one million units. It focuses on the AI Agent middle - platform and is installed on various hardware products such as elderly companion robots and smart walking sticks.
These new players in the market all share a common profile: They are small in scale, have limited backgrounds, and were not born with a silver spoon in their mouths.
So, the question arises: Why are these startups the ones? Meanwhile, facing sky - high traffic costs, how do they find their way to survive in the encirclement of giants?
Xiaoshui Intelligence may be the best case for observation.
01
How Does the Children's Smartwatch Achieve the Flywheel Effect?
To understand why Xiaoshui Intelligence has succeeded, we need to understand a basic fact: In many hardware markets, the children's smartwatch market is a typical example of a market with a huge gap between supply and demand.
From the perspective of the overall market, it seems to be a very attractive business. Public data shows that the global market size of children's smartwatches exceeded $5 billion in 2024. The data from Counterpoint Research is more intuitive: In 2024, the children's smartwatch market grew against the trend by 7%, while the overall smartwatch market declined by 4%.
In addition, in China, the world's largest consumer electronics market, the penetration rate of children's smartwatches is still increasing, especially in the sinking markets and third - to fifth - tier cities.
The "ice" comes from the competitive landscape. Companies like Xiaotiancai, Huawei, Xiaomi, and 360 are all around. Each of them has a mature supply chain, a large - scale distribution network, and a large marketing budget. Price wars break out every once in a while, but few players can create truly differentiated products.
Most manufacturers have fallen into a dangerous comfort zone: Relying on channel advantages and brand recognition to hold on to the existing market, they ignore the potential problems caused by the stagnation of product definition and the integration of new technologies.
As a product that accompanies children 24 hours a day, for parents, the core requirement of a children's smartwatch is to know where their children are at any time. Other functions are just icing on the cake. This safety anxiety drove the first wave of growth of children's smartwatches, but it stopped there.
Because the real users of the watch are children, and their core needs are fun: entertainment interaction, peer socializing, and exploring the world.
When the conflict between these two cannot be reconciled, most children lose interest in the watch after a few days and leave it in the drawer to gather dust. Industry data shows that the average monthly active users in the entire children's smartwatch industry are in single - digits, and this is based on the fact that some super - star products have call functions.
This is a huge structural pain point and also a huge but overlooked structural opportunity.
So, in early 2025, Xiaoshui Intelligence was the first to complete the private deployment of the DeepSeek large - language model and developed a vertical small model suitable for children's interaction, and launched an AI children's smartwatch that understands children better.
Based on the Deepseek model, Xiaoshui Intelligence has carried out a series of developments, turning it into a 24 - hour AI butler, a children's AI growth partner, a safety sentinel that can help parents monitor children's safety at any time, and a digital friend that can remember children's interests and learning progress. The popular intelligent agent "Mood Tree Hole" in the watch was well - received by children as soon as it was launched, with the highest single - user dialogue volume reaching 15,000 times per month.
In a previous public interview, Sun Xuefeng, the CEO of Xiaoshui Intelligence, recalled several important version iterations of the company's products:
"At the beginning, it was a voice - interactive watch, and children could talk to the watch. Later, we developed an AI picture - book function because it's difficult to answer questions like 'What does a Tyrannosaurus rex look like' through voice. The AI picture - book can directly generate pictures on the watch for children to see. Then came the video form, which could make dinosaurs move and make sounds. Now, the 4.0 version is an AI super - assistant for children, a long - term memory and growth AI partner that can solve learning and life problems for children."
However, making children love using the smartwatch through fun functions is just the surface. The deeper logic is that the preference data generated when a child interacts with the AI storytelling intelligent agent can be seamlessly transferred to the AI learning - tutoring intelligent agent, and the knowledge learned by the tutoring intelligent agent can be reviewed and strengthened in, for example, the science - popularization intelligent agent.
The data generated by children's voice conversations, Q&A, and stories several times, even dozens or hundreds of times a day, after being fed into the middle - platform, can further optimize the performance of the Agent. This is a scarce resource that no general large - model can replace.
So, the smarter the AI Agent is, the more users rely on it; the more users rely on it, the more data is generated; thus, a flywheel of user experience starts to spin.
02
How to Turn the Tough Hardware Business into an Attractive One?
Although the annual shipment of the smartwatch has exceeded one million units, Xiaoshui Intelligence does not position itself merely as an AI smartwatch company, or even just an AI hardware company.
Sun Xuefeng has emphasized on multiple occasions that Xiaoshui Intelligence is essentially an enterprise - level AI Agent ecological empowerment platform that outputs AI Agent middle - platform capabilities.
The fundamental reason is that the charging and business models of the two are completely different: The fate of hardware is a one - time transaction, while the subscription model is a business model that can support a company's stable and continuous cash - flow growth.
An early - stage investor recalled: In the past consumer electronics era, to make a profit from hardware, a company had to produce a device. However, each device was associated with a supply chain, yield rate, and inventory turnover, which involved a lot of dirty and tiring work. In terms of profit, cash - flow, or economies of scale, it was not very attractive.
The highly asset - heavy nature of hardware companies also limits their expansion. Correspondingly, the valuation logic of the capital market is also relatively conservative, with the PE multiple usually between 10 - 20 times.
However, after the emergence of large - models, the story has taken a new turn.
With the support of AI Agents, selling hardware is just the beginning. The AI agent can then continuously charge monthly fees through value - added subscriptions. This means that each sold piece of hardware will become an entry point for continuous cash - flow generation.
The essence of this model is to embed the Internet logic of "stickiness equals assets" into hardware products.
Let's do some calculations: Assuming that a subscribed user pays $20 per month (for basic AI services), even if there are only 500,000 paying users, the annual subscription revenue will be $120 million. This does not include the revenue from advanced services (such as AI computing power packages, character IP intelligent agents, and customized stories).
Once this story is well - told, the valuation logic of all smart hardware companies will also change fundamentally: The continuous paid - subscription module can be valued according to the PaaS/platform company model, evolving from a 10 - 20 times PE valuation to a 10 - 20 times PS multiple.
In fact, there are numerous successful cases of the hardware + subscription model globally.
The smart ring Oura is the most typical example. This small device, weighing only 3.3 grams, is sold at $349, and a monthly subscription fee of $5.99 is charged. In 2024, Oura's cumulative sales exceeded 2.5 million units, with a revenue of $500 million and a valuation of as high as $5.2 billion.
There are many similar stories: The AI card recorder Plaud Note, with a monthly subscription fee of $9.9, has an annualized revenue of $100 million; the intelligent pet tracker Tractive has an annualized subscription revenue of over $100 million...
In essence, they are all promoting an idea: The business model of AI hardware is AI, not hardware.
This is the second - layer logic of the flywheel, turning continuous interaction into continuous subscriptions and cash - flow, and transforming a hardware company into a platform company.
However, Xiaoshui Intelligence's ambition does not stop there.
03
Expanding in All Directions: How to Replicate Blockbusters?
In a gold rush, the most profitable ones are often those who sell water.
All hardware companies also have the ambition to transform from selling products to selling capabilities, and even selling solution standards.
Xiaoshui Intelligence is no exception. And to sell capabilities, it needs to first show the outside world that it has the ability to create blockbusters in batches.
In the past year or so, the AI war has almost turned into a battle of advertising investment. During the Spring Festival alone, ByteDance invested over $1 billion. Doubao was deeply involved in the CCTV Spring Festival Gala interaction, giving out hundreds of millions of red envelopes and gifts. Alibaba sponsored the Spring Festival Galas of four local TV stations and spent $3 billion on the Spring Festival promotion of its AI assistant Qianwen App. Tencent invested $1 billion, and Baidu invested $500 million to drive traffic to their AI applications.
The sharp increase in traffic costs has made life difficult for small and medium - sized hardware enterprises. At this time, the traditional offline channel business has seen a re - evaluation of its value.
And this is exactly what Xiaoshui Intelligence has been practicing hard for three years.
Since 2023, Xiaoshui Intelligence has continuously invested and accumulated channel resources of more than 17,000 offline communication stores across the country. This is an asset that has been seriously underestimated by the Internet circle and is the most ready - made distribution network for many small and medium - sized hardware enterprises that do not have the resources for online advertising investment.
Take the silver - haired economy, a new business area that Xiaoshui Intelligence has recently entered, as an example. Currently, China is entering an aging society at an alarming rate. It is estimated that by 2035, the population over 60 will exceed 400 million, accounting for more than 30% of the total population. In 2023, the market size of China's silver - haired economy reached about $700 billion, accounting for about 6% of GDP. It is estimated that it will exceed $1.2 trillion in 2025 and is expected to reach $3 trillion in 2035.
However, traditional elderly - friendly hardware faces an awkward reality: The elderly are not good at online shopping. To influence their minds, offline communication stores are the best choice.
In addition, traditional silver - haired intelligent devices are too primitive.
Whether it is a smart walking stick, a smart bracelet, or a health monitoring device, the underlying logic of these products is data collection. However, few people explore how to improve diet, control exercise volume, and remind medication based on the collected data. Not to mention how to add the most needed companionship function to these 24 - hour personal products.
Xiaoshui Intelligence saw this opportunity.
They installed AI Agents on products such as smart wheelchairs and smart walking sticks, upgrading the hardware to a considerate child for the elderly and an all - weather health sentinel, solving practical problems such as the elderly's loneliness, forgetfulness, and health monitoring.
For example, to accompany the elderly in chatting, or in response to the elderly's love for singing red songs, Xiaoshui Intelligence launched a red - song teaching Agent - the first few songs are taught for free, and after a certain number, a fee is charged according to the number of uses. A similar logic can be continuously expanded, such as a medication reminder Agent, an emergency call Agent, and a video reminder Agent for children.
Of course, the vision is not limited to this. The successful experience of the AI children's smartwatch can be replicated in elderly health devices; the solution for smart wheelchairs can be transferred to smart table lamps; the data insights from one scenario can nourish the product iteration in another scenario.
Xiaoshui Intelligence started with cloud communication, served the three major operators, cooperated with Hikvision and Dahua, and helped many enterprises in Huaqiangbei create AI blockbuster products. Now, with an annual shipment of one million children's smartwatches as a successful example, based on the AI Agent middle - platform, the company's AI capabilities are crossing from self - use to empowering partners.
This is a much larger market narrative.
In the past two decades, 80% - 90% of most smart hardware products, including children's smartwatches, have been produced in China, especially in Huaqiangbei. However, these hardware merchants with strong hardware manufacturing capabilities often have certain shortcomings in software capabilities.
A simple example is that to make an open - source model well - adapted to smart wearable devices in the cloud, it often needs to be distilled into a 7B - sized version, which will also lead to a sharp increase in the model's hallucination rate. This requires enterprises to be able to fine - tune the model or customize the knowledge base according to the needs of different scenarios. However, most enterprises do not have relevant capabilities at all.
What Xiaoshui Intelligence does is to modularly output the underlying middle - platform capabilities, providing the latest enterprise - level technology combinations, including end - side hardware execution, efficient AIGC generation, and operating - system - level intelligent collaboration and workflows, to help large, medium, and small enterprises fully automate complex creative, analytical, and execution work.
In other words, any company with hardware capabilities can privately deploy Xiaoshui Intelligence's AI agent middle - platform and quickly gain the ability to produce AI hardware.
And this is exactly the "terrifying" part of platform - type enterprises: Through an AI Agent middle - platform polished by millions of real - user interactions and a mature, low - cost offline channel, products can almost penetrate into any needed scenario.
With such a platform, in the AI era, Xiaoshui Intelligence can not only survive but also thrive with value.