From glasses to guitars, AI hardware is no longer obsessed with "new species".

In the AI era, hardware manufacturers are vying for the next entry point.

Although mobile phones and computers are still important traffic entrances at present, it is clearly observable that the rapid iteration of large models is pushing AI into a broader physical world. From smart glasses and AI earphones to smart rings and AI guitars, hardware manufacturers are all eager to separate AI from mobile phones, and even break away from the high dependence on Apps in the mobile Internet era.

From this, we have found an interesting rule: even if the carrier of AI is changing, the appearance of hardware still adopts forms that people are familiar with. Manufacturers are starting to use commonplace products to meet new intelligent demands. Behind this is not a simple choice of appearance, but a carefully weighed business calculation. Only when technological innovation enters real scenarios and forms the most direct connection with users' daily lives can it evolve from a technical Demo in the laboratory into a truly viable business.

So what kind of AI hardware can truly meet market demands?

Familiar scenarios are the business foundation with the lowest education cost

In the AI era, creating an unprecedented new hardware "species" seems to be more imaginative, but it also means higher market education costs.

A device that has never existed before needs to constantly answer questions like "What is it? What is it for? Why not use a mobile phone instead?" The more questions there are, the higher the user's understanding cost, and the greater the business uncertainty that follows.

Humane AI Pin once tried to create a brand new entry point outside mobile phones, but due to unclear usage scenarios, the product was eventually discontinued. Another well-received wearable device, Looki L1, has verified early sales of nearly 10,000 units through scenarios such as automatically recording life, generating Vlogs and life memories. However, compared with mature consumer electronics categories, it still needs to answer a longer-term question: apart from geek users who are willing to try new things, how many ordinary consumers are willing to wear a camera device on their chest for a long time?

In contrast, mature categories such as glasses, earphones, cameras and guitars already have relatively clear usage scenarios. Users know that glasses should be worn on the face, earphones are suitable for listening to music and making calls, cameras are used for recording and creation, and guitars are closely related to playing and singing, entertainment and social expression.

The market has given feedback on AI glasses first.

According to Omdia data, the global shipment of AI glasses reached 8.7 million units in 2025, and is expected to exceed 15 million units in 2026. China has become the fastest-growing AI glasses market in the world, with shipment volume second only to the United States.

Whether it is Ray-Ban Meta, which occupies the main share of the global market, or Rokid and Quark AI glasses in the Chinese market, the products must first look like a pair of glasses that consumers are willing to wear daily.

The business foundation of AI earphones is even more mature. Earphones themselves have already covered high-frequency scenarios such as music listening and commuting, while voice is the most natural AI interaction method. At the 2026 World Artificial Intelligence Conference, Qwen's first AI earphone made its official debut, integrating capabilities such as simultaneous interpretation and meeting minutes into the device that users already wear on a daily basis.

In more vertical categories such as cameras and musical instruments, the frequency of user use may not be equally high, but the corresponding interest demands, usage scenarios and payment habits have long existed.

Music is a typical scenario.

Consumers have long been willing to pay for devices such as earphones, speakers and musical instruments. Although musical instruments are not hardware that everyone uses every day, playing and singing, entertainment and social expression are demands that have been repeatedly verified by the market. The Tianpule AI Guitar released last year also unveiled its annual updated model at this year's World Artificial Intelligence Conference. According to previously disclosed data, the GMV of the product exceeded 10 million yuan less than three months after its launch. This result is not enough to prove that AI guitars have become a mass category, but it shows that users are willing to pay for a new music interactive experience.

At first glance, it looks very much like a stringless guitar with a screen attached. Jia Shuo, Vice President of Zhuitun Technology, has no intention of correcting consumers' understanding. In his view, if consumers think a product is significantly different from what they have seen in the past and belongs to a brand new category or species, it often means extremely high market education costs and will also bring great business uncertainty. "Consumers are always right. What they think it is at first glance is probably what it is," Jia Shuo believes.

However, when dismantled from the product perspective, the internal logic of this product has gone beyond traditional stringless guitars. To embed the music generation large model into it, Tianpule has made a large number of modifications at both the hardware and software levels, integrating a screen, speaker, interaction system and AI music generation capabilities. To a certain extent, it is closer to an "all-in-one AI music large model device", but the first product chose the familiar guitar shape that people are used to.

Nevertheless, familiar scenarios and forms only lower the first threshold for AI hardware to enter the market. The next step that determines the value of AI hardware is whether AI can enter the core usage process of the product.

Connected to AI does not mean being AI hardware

In the boom of AI hardware, accessing a large model, adding voice Q&A, or making hardware connect to mobile Apps are no longer difficult things. But if AI only exists outside the hardware and the original product logic has not changed, it is still difficult to form a new reason for purchase.

The evolution of smart guitars provides an intuitive sample.

Most early smart guitars adopted the "traditional guitar plus external attachment" mode: adding a third-party auxiliary key accessory to the guitar to help users quickly master guitar chords, which is more like a set of auxiliary teaching tools.

The stringless guitars that emerged later made drastic changes to the inherent form of guitars. They canceled traditional strings and interacted through buttons, touch and other methods. Stringless guitars are more like a guitar-shaped audio sampling player. After users connect to the mobile App, they can quickly complete playing and singing by pressing keys according to light prompts.

These attempts lowered the learning threshold of musical instruments and made more novice users willing to pick up a guitar. However, mobile phones still undertake core functions such as content selection, music score display and interactive control, and the hardware is more of an extension of the App in the physical world.

Similar problems also exist in other AI hardware.

If AI glasses only move the voice assistant in the mobile phone to the temple, it is still more like a hands-free mobile phone accessory; if AI earphones are only responsible for sound collection and playback, their AI attributes mainly stay at the interface level.

The real watershed lies in whether AI has changed the core experience of users.

A key change of Tianpule AI Guitar is that it integrates the screen, software and music model into the guitar body, so that music scores, lyrics, accompaniment and AI creation no longer rely heavily on mobile phones. When playing, users do not need to look at the guitar neck while looking for the mobile phone placed next to them, nor do they need to switch repeatedly between multiple devices.

"The process of playing music requires more immersion and focus," said Jia Shuo. If a device can fully present all the information required for user operations, it can reduce the interruption of the mobile phone to the playing process, making the product closer to a smart musical instrument that can be used independently.

This is also one of his criteria for judging AI hardware: "A truly smart hardware in the AI era should be able to deliver a complete intelligent experience even without the App."

The hardware is no longer just a remote control for the mobile App, but has begun to become a terminal that can independently deliver core experiences.

At this point, the changes brought by AI are no longer limited to efficiency improvement, but advance the creation and positive feedback.

AI guitars not only help users "play a song by someone else", but also allow users to generate music according to their own emotions and preferences, and then participate in performance and expression. From "quickly learning a song" to "quickly getting a complete music experience", what AI truly lowers is not only the technical threshold for playing music, but also the threshold for ordinary people to gain a sense of achievement through music.

This change is not exclusive to musical instruments. In imaging equipment, DJI and Insta360 use capabilities such as automatic following, stabilization and intelligent camera movement to allow ordinary users to obtain high-quality images faster without mastering professional shooting skills first.

However, lowering the threshold does not mean simplifying the product infinitely.

In Jia Shuo's view, if a music hardware only allows users to press a few buttons and play fixed accompaniment, although it is easy to get started, it may quickly lose its appeal due to lack of challenges and sense of participation.

This is also the balance that Jia Shuo repeatedly mentions, which is commonly faced by consumer-grade AI hardware: it must be easy enough for first-time use, and also allow users to obtain higher-quality results and have space for further exploration after they are familiar with the product.

The product logic of DJI Pocket is a typical example. Its power-on, shooting and zoom operations are intuitive enough for novices to use quickly, while its output quality can also be accepted by professional studios. The product does not stack complex operations for professional capabilities, nor does it sacrifice the final result for simplicity.

AI is rewriting the product cycle of hardware

Over the past decade, "software-defined hardware" has become an important paradigm in the consumer electronics industry. From small devices like Bluetooth speakers to large products like new energy vehicles, software is continuously changing the functional boundaries and interactive experience of hardware. Nowadays, traditional software is being swallowed and reconstructed by AI, and large models have begun to participate in the product definition of hardware. Jia Shuo summarized this change as "model-defined hardware" at last year's Yunqi Conference.

When models begin to participate in product definition, the life cycle of hardware will also change accordingly.

In the past, the capabilities of traditional consumer hardware were basically fixed when it was launched on the market, and new functions that truly affect the experience often had to wait for the next generation of products. AI hardware is more like a terminal that can continue to grow.

After the Tianpule AI Guitar was launched, it basically maintained an update rhythm of once every two to three weeks; in the more than half a year since its launch, the core gameplay has been adjusted two or three rounds. The newly added "AI Improvisation" function this year is open to all users for free through firmware upgrade.

"This is a muscle memory formed by mobile Internet teams in the development process," Jia Shuo joked.

This iteration is also extending from software to hardware. To weaken the toy-like feeling of the product and reserve space for advanced users to continue exploring, the 2026 annual updated model has redesigned the rainbow string component to further improve sensitivity and playing experience.

The competition methods of AI hardware companies are also changing accordingly.

Tianpule's journey from App, large model, Agent to hardware did not happen overnight. As early as 2019, the team took the lead in defining the prompt light "playing and singing" interaction through the Changya App, which to some extent inspired the smart musical instrument products on the market in recent years. After that, the team began to independently develop the Tianpule music large model, and incubated Tunee, the first conversational music creation Agent in China. On this basis, the team further pondered: how can the music generated in mobile phones and computers truly enter users' daily life and social scenarios? In the end, hardware became the centralized outlet connecting models, software capabilities and existing demands for playing and singing.

This path may not be applicable to all companies, but it reveals an increasingly clear trend in AI hardware competition: whether the product can continue to evolve after launch, and turn short-lived freshness into long-term value in repeated real usage.

The combination of AI and hardware is still in its early stage. AI itself is still iterating rapidly. How to stably convert the ever-changing model capabilities into hardware experience will be a long-term topic.

From glasses and earphones to guitars, AI hardware no longer indulges in "new species", but re-enters those existing scenarios, solves problems that have not been properly solved in the past, and creates new experiences.

What will determine the success or failure of the industry in the future will not be who first installs AI in a product, but who can truly connect models, software and hardware, and turn technology into a natural, stable and long-term usable experience.

When the freshness fades away, users are still willing to keep using it, which is the real beginning of a viable AI hardware product.