The AI glasses market is heating up globally, but industry veterans warn that the current frenzy is more about expectations than guaranteed success, emphasizing that long-term growth will depend on solid product experience and strategic focus rather than sheer publicity.

Yang Longsheng, CEO of INMO, said the industry is “on fire” this year as lightweight, all-in-one AI glasses capture market attention.

He shared his insights with Wu Dezhou, founder and CEO of BleeqUp; and Wan Ning, Co-founder of TMTpost Group, Founding Director of ITValue, and Director of TMTpost Research Institute, during a roundtable discussion themed "The Buzz and the Harsh Reality Behind the Boom in ‘Smart Hardware’ Startups" on Wednesday during the 2025 T-EDGE conference, which kicked off on Monday, December 8, and runs through December 21. The annual event brings togehter top scientists, entrepreneurs and investors to discuss pressing issues of the AI era.

“Previously, when people thought of smart glasses, they imagined bulky VR headsets or devices like Vision Pro. Since the end of last year, the focus has shifted dramatically to AI glasses,” Yang said.

He credited the initial surge to Meta’s Ray-Ban collaboration, with Xiaomi, Alibaba, and other tech giants further accelerating the category’s maturation.

According to IDC, global shipments of smart glasses reached 4.296 million units in the third quarter of 2025, a 74.1 percent increase year-on-year. Yet, Yang cautioned that market attention does not automatically translate into sustainable success.

Wu described the current excitement as an early-stage “expectation bubble” typical of emerging smart hardware sectors.

“Aligning marketed features with actual product experience is the cornerstone of building long-term user patience,” Wu said.

He emphasized that true competition lies in creating solid experiential barriers in vertical use cases, requiring focus, restraint, and long-term product commitment beyond short-term hype.

Yang added that AI glasses are one of the few innovative hardware categories with relatively low speculative risk. “Vertical scenarios are already validated, and the products are useful to target users today and will be even more useful tomorrow,” he said.

Both Wu and Yang are veterans of China’s smartphone industry who view AI glasses as a potential “super terminal” that could replace smartphones. However, the entry of tech giants inevitably reshuffles the market. When Alibaba launched its Quark AI glasses, the company’s massive publicity temporarily overshadowed smaller startups.

“Facing giants, startups must avoid direct competition and instead focus deeply on vertical scenarios with high demand,” Wu said, describing a “needlepoint strategy” similar to approaches used at Huawei.

Yang noted that Xiaomi’s market entry initially tripled e-commerce traffic for their products, with Alibaba’s entry later boosting it eightfold. “This forces us to benchmark against top-tier manufacturers and improve user experience, or we simply won’t survive,” he said.

IDC forecasts that shipments of smart glasses in China will reach 4.508 million units in 2026, a 77.7 percent increase, while AR/VR devices are expected to grow 62.1 percent year-on-year. Yet AI/AR glasses remain niche compared with smartphones and PCs.

Looking ahead, both Wu and Yang predict AI glasses could become mainstream consumer products within two to three years, while AR glasses may take three to five years or longer. “Smart glasses are now recognized as the next-generation terminal device,” Yang said. “The biggest challenge for startups is pace: move too fast or too slow, and you risk losing out.”

The transcript of the roundtable discussion is as follows, edited for clarity and brevity:

Wan: With the surging interest in AI heading into 2025, how do you both view the current market? What new perspectives do you see from capital and users compared with previous years? And how does this affect your company’s development?

Yang: Our company has been focused on the smart glasses category for five years now, dedicated to lightweight, all-in-one AI+AR eyewear devices. Today, we’ve taken the lead in the domestic lightweight segment. Moving forward, our unwavering mission is to make glasses the next-generation mobile terminal replacing smartphones.

To answer your previous question, my biggest impression this year is that the industry has truly taken off. In previous years, when people mentioned smart glasses, they would think of bulky VR headsets or products like the Vision Pro. Since the end of last year, though, there has been a dramatic shift: market focus has landed squarely on lightweight, integrated AI glasses. This surge in popularity was initially sparked by the Meta Ray-Ban, and then major internet companies like Xiaomi and Alibaba joined in, propelling the entire category into the spotlight. For us, a more crowded track certainly brings challenges, but having the entire sector under the spotlight is definitely a positive for the industry.

On the other hand, starting this year, more and more mainstream users have begun to pay attention to the smart glasses category. Ordinary users are now proactively asking, “How do you choose AI glasses?” A couple of years ago, when I started working on this, my friends had no idea what I was doing; but now, they compare different brands like Quark Glasses and are even ready to purchase directly. Public awareness has essentially been “educated” all at once—this is the biggest benefit of this year. Looking deeper, what AI brings to glasses is no longer just photography, but a whole set of scenario extensions—the market is truly being opened up in this way.

Wu: Our primary focus is on smart glasses for outdoor sports scenarios. This year is being called the “Hundred Glasses War,” with many brands focusing on general-purpose AI glasses, but we’ve chosen to go down a vertical track—specifically sports glasses. We aim to solve scenario-specific pain points for over one billion outdoor sports users worldwide and use AI to reshape the sports experience. That’s what we’re working on.

As for the state of AI this year, as Mr. Yang just mentioned, everyone can sense it—2025 could very well be the first year for AI hardware. Right now, people are paying a lot of attention to AI hardware and have tried many products, including devices like the AI Pin. But what I’ve found is that there’s a growing consensus: glasses may become the best hardware carrier for AI in the future and the next gateway for user interactions. That’s why you’re seeing more and more major internet and smartphone companies entering the smart glasses market in the second half of this year. I think this is a great thing—it shows that glasses are moving from being novelty gadgets for tech enthusiasts to entering a new stage where they play a strategic role in the ecosystem.

From the user’s perspective, we can clearly sense the difference. Although many people are discussing, buying, and experiencing smart glasses, everyone is forming their own opinions. Some feel the products meet their needs, while others voice quite a few complaints and suggestions. I think all of this is very positive. Just like in the era of smartphones, only when users have actually used the product can they truly know what they want and what problems they hope the product will solve.

Wan: There’s a difference between “someone using it” and “something becoming a long-term hit.” Could this just be the opening ceremony of a new bubble? After all, both your companies received major funding and a lot of attention in the past rounds—how do you see the current situation?

Yang: I think this is a rather sensitive question. In my opinion, compared to many other innovative products, smart glasses are actually one of the categories with the least hype. At present, specific vertical scenarios have already been proven—for example, the sports glasses developed by Mr. Wu’s team, or our recently released GO 3 for real-time translation—users are willing to pay for concrete functions. This isn’t just storytelling; it’s about real utility. If a product solves a problem and people are willing to buy it, then it’s not a bubble.

Going forward, the evolution of smart glasses will increasingly align with smartphones. Both Mr. Wu and I come from the smartphone industry, so we have a clear perspective: after one or two more iterations, smart glasses will gradually shift from being a "tool" to having an "entertainment" attribute. Today they’re used for navigation and translation; tomorrow, users will be able to scroll TikTok or watch short videos—the feature set is steadily moving closer to that of smartphones. Recently, Google officially released its XR system, and with the popularity of Doubao AI, the groundwork for interactivity and ecosystem on the glasses side has essentially been laid. This trend is irreversible and advancing even faster than we imagined.

In another two or three years, all these capabilities will be fully integrated into smart glasses, transforming them from mere tools into everyday entertainment devices, and even social terminals. By then, they will undoubtedly become the "next-generation smartphone," delivering scenarios and functions that smartphones simply can’t. This is the path we’ve always believed in and remained committed to. So to sum up: smart glasses are already "useful" to vertical users today and will be even "more useful" tomorrow. This industry is definitely not a bubble.

Wan: Mr. Wu, do you agree with Mr. Yang’s assessment? He also emphasized our deep engagement in vertical scenarios. What are your thoughts on the so-called "bubble"?

Wu: How do you define a bubble? In my view, the AI glasses industry is only just getting started. It’s precisely because the intelligence of large models is advancing so rapidly that glasses can truly address user pain points. If you compare the entire industry to smartphones, we’re at the explosive early stage—the pace of development is fast, and there’s a lot of attention, which might give people the impression of a bubble.

But there are indeed some current issues, such as relatively high return rates or certain user experience complaints. In fact, this can be understood as a kind of "expectation bubble." There is still a gap between what users hope for in terms of features and the actual experience with the product. After living through the mobile internet era, expectations for smartphone experiences have become extremely high, so naturally people have high hopes for smart hardware too, imagining that it could solve many problems. However, when they finally get the product, the experience still isn't perfect, and this gap makes people feel like the product was overhyped and underdelivered. This is something every smart hardware industry goes through in its early stages. In fact, I hope everyone can be a bit more patient, and also call on the industry as a whole to promote products more rationally, aligning the features presented with the real experience. That way, users will have more patience to wait for product iterations.

Wasn't it the same when smartphones first came out? The cameras were only 300,000 pixels, and now they're approaching DSLR quality, becoming the main photography tool in daily life. Smart glasses have weight limitations—it’s natural that they also need an iteration process. Criticism is normal; at the very least, it shows that people have truly used the product. And if you look at Meta, it sold 2 million devices last year, and it’s said that shipments will exceed 10 million this year. In fact, a large number of users are already accepting the current experience and starting to use it to solve real-life problems. Of course, we still have many areas to improve, but this is a phase the industry inevitably has to go through—in my view, that's actually a good thing.

Wan: What impact do you think the entry of tech giants will have? Will you adjust your product strategy and direction because of it?

Wu: At the beginning of last year, we already made the judgment that all smartphone manufacturers would eventually develop AI glasses. This is going to become their main track. In the short term, smartphones won’t be replaced, but a lot of daily interactions really don’t require taking your phone out anymore—input and output can be done directly through the glasses. This is likely the main reason we see everyone entering this field together.

The arrival of tech giants is an enormous positive for the entire industry. With their vast resources, ecosystems, brand influence, and capital, they're able to educate both consumers and the industry more efficiently, quickly lower the barrier to understanding, and grow the overall market. It would be very difficult for startups to accomplish this market education on their own in such a short period of time.

That said, the entry of tech giants also brings challenges to startups. So, we've chosen to avoid direct competition and focus on more vertical, high-demand scenarios, deeply cultivating those areas. This is a choice we've made based on our judgment of trends.

Yang: That's right, this year we’re facing direct competition—it’s a true head-to-head battleground. I could clearly feel that since the second half of the year, with both Xiaomi and Alibaba entering the market one after the other, the overall industry traffic has become extremely volatile. When Xiaomi just came in, e-commerce traffic immediately tripled; then by the end of the year, when Alibaba joined, traffic spiked another eightfold—the growth was dramatic. Take companies like Yingmu, for example—who have been deeply cultivating this track for four or five years—they only just released their latest AI glasses this year. The pair I’m wearing now, the GO 3, looks just like regular glasses—very everyday and low-key.

So we immediately felt the shifts in the market: as the tide rises, all the boats are lifted. Our overall sales have increased two to three times compared to last year. But as Mr. Wan just said, pressure followed close behind. In the past few years, we were like 'trailblazers' pushing forward, but this year, several major players all released similarly defined products, and all concentrated their launches in the second half of the year, which certainly hit us hard.

The pressure comes mainly from two fronts: First, the small user experience issues we didn't address well enough before are now being amplified. If you search on Xiaohongshu or Bilibili, you’ll find users tearing down our products and Alibaba’s one by one for comparison—every detail we previously overlooked is now being put under the microscope. This forces us to match up to the standards of top tech giants and deliver an experience even better than theirs—otherwise, we simply won’t survive.

From another perspective, the big internet companies position themselves very differently from us. Alibaba sees its Quark Glasses as an extension of its ecosystem—just a new carrier for its AI; but from day one, we’ve treated glasses as an independent device, not reliant on anyone else. So our choices for ecosystem and technical path are totally different from theirs. We dig down to the optical and chip source design ourselves—even our OS is self-developed. At the same time, we open up our ecosystem, partnering with Tencent, Zhipu, and others to build out the content—our goal is to make glasses that can both run independently and remain open for other integrations.

Our core objective is to make this a device on the level of a smartphone—not just an accessory or extension of an existing ecosystem. This strategic choice might be the crucial way for startups like us to truly survive and carve out independent space in a landscape dominated by industry giants. After all, what we want to accomplish is something deeper and more long-term—not just short-term market opportunism or simply stacking up features. This positioning reflects our deep thinking about the future of the industry and a long-term commitment to user value. It also means we’re determined to take a path that’s truly our own.

Wan:I’d like to have Mr. Yang continue: Compared to industry giants, your team is focused, highly specialized, and you’re driving end-to-end innovation—covering optics, chips, and hardware experience all at once—almost like a “special ops” strategy. How do you differentiate yourself from the giants and carve out your own space in the market? Ultimately, how do you ensure the industry flourishes with a diversity of players, instead of just being dominated by a few?

Yang: Simply put, if we want to turn AI glasses into the next-generation smart device, we have to invest for the long haul—hardware, OS, supply chain, ecosystem—none of these can be overlooked. In the past, only phone manufacturers operated this way. Take Huawei, for example—a single product might take eight or ten years of preparation. Now, if glasses are going to take over, we have to invest on the same timeline.

Since the company was founded four or five years ago, we’ve gone all in on the “integrated AI+AR glasses” track. We built our own optics manufacturing line, assembled an in-house OS and algorithm team, and then partnered with major internet companies to develop content together. Every step burns cash, every step takes five or ten years of persistence before glasses have any chance of becoming the next-generation smartphone. This kind of long-term, heavy investment is something you won’t necessarily see from the major internet companies.

The giants usually partner with hardware companies like GoerTek, then layer on specialized content themselves, treating smart glasses as an extension of their closed ecosystems—a model that is faster and requires less depth. But if you zoom out to a five- or ten-year horizon, the companies that survive will be those that have tackled the entire value chain—from core hardware and OS down to content ecosystems. It will never be those who just slap their brand on an ODM product. The smartphone era has shown us: success requires heavy, in-depth investment. That’s our biggest card to play in this differentiated competition with internet giants.

Wan:Next, I’d like to ask Mr. Wu: Mr. Yang is choosing to face the big players head-on since this field truly requires full-stack investment, from the underlying technology all the way to the ecosystem. But your strategy is different. You also come from a mobile background, yet you’ve chosen a different path. The three of us have just agreed: the market is big enough, the potential use cases are vast, and investors, consumers, and even users on Xiaohongshu are all brainstorming and exploring creative applications. How do you view this stage of the industry? And how do you determine your future strategic direction?

Wu: I really admire Mr. Yang’s choice. The general market is indeed larger—overseas, the average user has five or six pairs of glasses, and in the future, the number of glasses will definitely surpass smartphones. But our strategy is not to compete head-to-head with the giants. Instead, we focus more on the vertical sports market.

This aligns with the “needle-point strategy” we insisted on during the Huawei era: when resources are limited, concentrate on one area and go deep enough, get thorough enough. Our company’s goal is to become the Garmin of the AI era, focusing exclusively on sports scenarios. At present, we’re entering through cycling, running, and trail running, and plan to expand into activities like skiing and more sports scenes in the future.

However, sports scenarios are quite different from general consumer scenarios. General consumer products require heavy investment in brand marketing, launching new products, and rapid mass communication; whereas sports brands like Lululemon and Shokz typically start by targeting professional users. They rely on authentic user experiences and content created by sports enthusiasts, building up word-of-mouth credibility within private communities. Therefore, our primary marketing strategies revolve around sponsoring events, participating in niche expos, and enabling key influencers to experience our products and generate content.

Zhu Xiaohu once said that compared to large companies, this might be exceptionally tough and thankless work. But we want to fully immerse ourselves in these tough, hands-on tasks and put significant effort into user operations. Whether through official channels or private communities, we genuinely listen to our users’ needs and solve the pain points they encounter in sports scenarios.

Mr. Wan is a sports enthusiast himself and needs to carry many devices when exercising: sports glasses, headphones, cameras, walkie-talkies, etc. Our first step is to integrate four devices into one piece of hardware, making it much more convenient to carry. While this might not be an absolutely essential need right now, at the very least it addresses the issue of convenience.

Beyond that, in terms of applications and services, we hope to create features that are uniquely suited to glasses. Since glasses are worn on the face and sit close to the senses, they can incorporate more sensors to capture biometric data. Looking ahead, we plan to offer value-added services focusing on sports health and safety. As for recording, we’ve already developed a one-click editing feature that uses sensors and large AI models to accurately identify highlights, rapidly generating shareable videos. These are pain points that other devices struggle to resolve in sports contexts, and we aim to build deep expertise around them.

Wan: I agree with your "pinpoint" strategy of building momentum in niche markets, but users nowadays expect more than just data from their wearables—what they really want is real-time advice that enhances value. Could you share your thoughts and plans for combining AI and smart technology?

Wu: While many functions can already be fulfilled by existing devices, what we care about most is the unique experience that AI can deliver—that’s the real core. If we can’t create an experience that’s superior to current products, it’s not meaningful. Sports users are also more loyal than general users, and there’s even a sort of hierarchy—people who play golf, for example, won’t wear Nike or Adidas, but will only choose specialized golf brands.

We focus on two main areas: first, capturing and sharing social highlights. The highlight moments vary in every sport. Previously, sports cameras relied on computer vision or sensors to identify highlight clips, but they couldn’t always capture every key moment. We hope to use large AI models to accurately recognize the highlight moments in videos for each type of sport—whether it’s cycling, fishing, or golf.

Secondly, with tools like Jianying, you can quickly generate videos with just one click. At the core, what sports users really need is health and safety while exercising. Our goal is not just to make camera glasses; we also want to integrate more sensor devices (whether developed by ourselves or others). We’ve already integrated with products like the Apple Watch.

Compared to other devices, glasses naturally offer both audio and visual input, providing a richer set of modalities. This allows for a more accurate assessment of a user's activity and state, enabling us to offer health and safety alerts during and after exercise. In the future, these could be offered as value-added services. Currently, major smartphone brands and internet giants are focusing on generic products, and they're not as deeply invested in sports scenarios as we are. In this era where major players are ramping up efforts to educate the market, the breakthrough strategy for startups is to build an ‘experience gap’ in vertical use cases, injecting reliable, sustained progress into the industry with solid product logic.

Wan: Now I’ll throw the question back to Mr. Yang. Since you’ve chosen to go head-to-head in the smart glasses market—and considering glasses are a high-frequency, close-to-the-eye wearable, with people's needs continuously amplified in the shift from informationization to digitalization, and with the external digital environment getting more sophisticated and new smart devices driving new scenarios—I’d like to dig deeper and understand: What is the core philosophy behind Yingmu’s current products, and how do you bring this idea of ‘professional depth’ into your design process?

Yang: If I had to sum it up in one word, it would be ‘lightweight all-in-one’. Five years ago, when we first introduced this concept, the market was filled with VR headsets or Birdbath (coaxial optical waveguide) tethered glasses—no one wanted to tackle this tough problem. We started by focusing relentlessly on arrayed optical waveguides, investing four or five years. It wasn’t until Meta’s launch last year that people generally agreed this is the lightest full-color solution. The OS followed the same principle—compressing the algorithms so they could run on an MCU in six layers; only with a lightweight enough architecture can you actually wear it daily, and look good doing so. Every bit of our tech investment centers around this idea.

We break the product down into three phases. In the current phase, it’s all about ‘toolification’, which echoes Mr. Wu’s view—first solve a single function, a specific scenario, and serve a particular user group. The X Series and GO Series focus on business travel translation: precise tools with a lifecycle of about one to two years. These won’t disappear, but will be upgraded in the next phase. The second phase will definitely feature full color, targeting entertainment: binge-watching shows, watching movies, gaming—the AIR Series is our attempt at this. Once the AIR Series becomes thin and light enough, it can directly replace the GO Series and other tool-type products; after all, if you can scroll Douyin on it, who’d use it just to listen to music? The final stage will arrive naturally after that.

Looking ahead, what I’m most optimistic about is still the social aspect—this is the ultimate sticky scenario for smart glasses. Simply put, when I’m wearing the glasses and see Professor Wan, the system knows you’re a friend of a friend and quietly prompts me, “Would you like to get to know each other?” Or if I glance at a girl, the AI analyzes and suggests, “She could be your future partner.” This kind of real-life ice-breaking is simply impossible to do in the smartphone era. The slogan could be “Don’t miss your chance with the right people around you.” We’ve been exploring this with partners, but in the short term, we haven’t yet found the right product that can fully support this scenario.

Back to the technical side, we’re relentlessly focused on “l(fā)ightweight integration.” Our product roadmap is clear: start with tools—translators and voice recorders—progress to watches and tablets, and ultimately achieve fully integrated smartphone-glasses. This is the roadmap we’ve consistently adhered to.

Wan: Facing a “professional + consumer” dual-track strategy, how does your management team allocate resources and prioritize between them?

Yang: Our view is that both the mass market and niche professional markets will coexist in the future, just at different volumes. The mass market could reach the scale of smartphones—with annual sales in the billions—while segments like sports and education could easily sell tens of millions or even hundreds of millions of units. I think everyone agrees on that. The difference is, we believe there won’t really be such a thing as “professional hardware” in the future. Hardware will converge, and the differentiation will be left to application companies. In simple terms, we’ve laid a “l(fā)ightweight integrated” technical foundation—from optics to OS—creating a hardware base that’s flexible and reusable.

So our approach is slightly different from Mr. Wu’s. We basically don’t do applications ourselves; instead, we bring in all sorts of partners to handle that. Previously, sports brands have approached us to tweak our hardware for use in golf, skiing, and other such scenarios—we simply didn’t put our own brand on those. What we focus on is deeply sharpening our vertical technology, making our foundational hardware robust enough to go head-to-head with Apple and Huawei. As for the horizontal breadth, we leave that to our partners to test out all kinds of industry-specific scenarios—they handle the content, we just provide the hardware. This is the path we insist on taking; every company has its own vision, and for us, we’ll keep digging deeper in this direction.

Wan: Right now, INMO Tech is starting to look a lot like Siemens in the electrical era—you want to build the largest universal foundation for applications, serving as the underlying platform that partners from various fields can use to collaboratively create an entire ecosystem. Today’s discussion is fascinating: the two of you take different paths—one is going for the ultimate universal hardware platform, while the other is perfecting the experience in specific scenarios. When your ideas collide like this, could it spark the chance to create something as globally unique as Garmin watches? Mr. Wu, what are your thoughts?

Wu: I’ve been asked this question many times: will smartphone companies also head in this direction in the future? My take is this—let’s use smartwatches as an example. Brands like Huawei, Xiaomi, and Apple all offer watches for sports scenarios, with heart rate monitoring and step counting. But real sports enthusiasts mostly choose Garmin, Coros, or Suunto. The difference comes down to the fact that universal products need to meet too many demands—the real issue is how to make trade-offs.

Especially when it comes to watches or, in the future, smart glasses, there simply isn’t enough space or flexibility to cover everything. It’s impossible to accommodate every user and every scenario. Universal products can only distill the common needs—like heart rate and step tracking—catering to beginners or light users. But professional users want more accurate GPS tracking, offline navigation, and longer battery life—not something that needs to be charged every day.

Take Garmin for example: from the beginning, they used a semi-reflective, semi-transmissive display (like old Nokia phones), which gets clearer under sunlight and saves a lot of power. That was their trade-off—sacrificing bright colors in exchange for battery life. For trail running, Garmin watches even have a built-in flashlight. Huawei and Apple might see that as neither stylish nor smart and just leave it out, but for trail runners, it’s a must-have.

The key is this: when you focus on a single scenario, it’s much easier to make trade-offs. We’ve worked on universal scenarios before, but with too many demands, it’s hard to prioritize, so you end up just adding more features, and none of them are truly refined. That’s why our approach is different from Mr. Yang’s—we choose to focus on sports.

We work to understand the essential needs of specific sports communities and solve real problems. In the future, there will be glasses for cycling, golf, and skiing—the basic functions (photography, audio, intercom) will be the same, but the backend services will be tailored to meet the unique needs of each sports group.

Here’s another point: unlike smartphones, glasses are fashionable items you wear on your face. In the future, brand concentration won’t be as high as it is with phones, where only around ten brands dominate. If you look at the eyewear and sunglasses market today, even the top ten brands rarely have more than a 10% share each—it’s highly fragmented. After smart glasses become mainstream, there will be fewer brands, but still far more than there are for smartphones, because fashion items must meet everyone’s individual preferences.

Wan: When your product positioning is to "focus on building great hardware as a foundation," you inevitably face the challenge of establishing interfaces with all partners. This places high demands on your business model. At the same time, you’re surrounded and pressured by big companies. How should we approach this?

Yang: You could say that Yingmu is the most open company in the smart glasses industry. We are very disciplined about our boundaries: we only do fundamental hardware and OS, leaving all content to our partners. There are several collaboration models. The simpler approach is integrating platforms like Tencent’s App Store or Alipay to jointly test different use cases. For deeper collaborations, we do co-branded products—where the partner defines the product, including weight, display solution, and app, while we focus only on the hardware and OS segment. As we don't have the brand baggage of major companies, as long as the scenario is valuable and expected to scale, we’re happy to collaborate.

Wan: I previously noticed that another competitor, Oakley, released the Vanguard, which is quite similar in form and use case to BleeqUp. Some people consider you a strong alternative to them. What do you think of this market positioning? Given that China’s manufacturing supply chain is second to none globally, was that an important reason for you choosing this track?

Wu: We are actually the first smart glasses in the sports field, but later, everyone started paying attention to the Oakley and Meta collaboration—the Vanguard—which may launch in October, while we’re already selling globally as of September. After Vanguard came out, we had many channel partners do side-by-side comparisons, and the feedback was straightforward: If a user cares primarily about the brand, they’ll certainly buy Vanguard, since Oakley has deep roots in the sports scene. But for actual cycling use, most think our BleeqUp is the superior solution.

The core feedback centers on two points: long-duration recording (similar to a dash cam), and the intercom function—these are essential needs for cyclists, while Vanguard lacks them. Vanguard can only record for three minutes at a time, with about half an hour of continuous shooting. We support 1080P recording for an hour straight, or up to five hours with a 10-gram external battery. Outdoor cameras are rare among cycling users in Europe and the U.S., so features for safety and evidence-keeping are crucial—this is the real, essential demand.

This shows that Oakley and Meta, as major companies, approach the market with a strategy of making general-purpose sports glasses first, leveraging brand recognition to attract users. Our approach is the opposite—even when entering the sports segment, we are committed to delving deeply into each specific sports scenario. By offering a comprehensive solution of hardware, software, and services, we strive to deliver the ultimate user experience. This way, even if other players enter the market, our experience in the cycling segment will already be the best; whenever users think of smart cycling glasses, they will think of us. We establish our foothold first, and then expand into other sports fields.

In addition, there are other product differences: the Vanguard weighs 66 grams, while ours is under 50 grams (48-49 grams); the Vanguard cannot change lenses, so it’s unusable for nearsighted users, whereas ours allows for easy lens replacement and even offers highly customized prescription lenses. These factors enable us to be more professional, more thorough, and more in-depth in our approach.

Wan:Just now, Mr. Yang mentioned that when it comes to social features, it usually takes a while to explain. I’d like to give you some time now—could you elaborate a bit? When glasses become a two-way channel for ‘how I see the world, and how the world sees me,’ capable of overlaying real-time social cues in ways smartphones simply can’t, what’s your vision?

Yang: This scenario has been at the heart of my entrepreneurial motivation—it's something I've been imagining for over a decade. Social interaction in the smartphone era is actually quite insular: you send a message and have no idea if the other person is male, female, or how old they are; people remain separated from each other.

Speaking from personal experience—many people born in the 90s, 00s, and even 05s actually long for the kind of offline conversations that were typical in your youth, Mr. Wan, like ‘chatting in a small courtyard while fanning yourself.’ The mobile internet has stripped away so much of that warmth from face-to-face interactions. What smart glasses need to restore is precisely that sense of ‘connection and warmth’—this is a fundamental need. The essential point here is ‘breaking the ice.’ With life moving so fast, people no longer have the luxury of a cozy courtyard for leisurely chats; there needs to be a reason on the spot for you to want to talk to someone on the street. Glasses are naturally able to provide this reason: as you pass someone, social tags pop up directly in your field of view—whether they’re open to chat, if they’re available, whether they’re interested in talking about work, romance, or hobbies—it’s all immediately clear, and the ice is broken in an instant.

In the future, offline connections can form just as quickly as they did for the older generation, and this is an irrepressible desire among young people. Glasses could become the stickiest scenario, but to make this happen we can't rely solely on hardware manufacturers; it will require content providers, data companies, and even government participation to truly build this ‘offline social network.’ That’s the more idealistic vision I have!

Wan:So Mr. Wu, I’d like to turn to you: What do you think of the future that Mr. Yang just described? Let’s broaden our horizons and look further ahead.

Wu: I still firmly believe that glasses will be the smart assistants of the future, acting as the physical carriers for your future Agent. Social interaction is just one part of this, especially as large language models become more powerful. Glasses are the most suitable input and output device—they offer multi-modal input like visuals and sound, letting large models understand your environment, actions, and language. They can connect with a broader ecosystem to bring you more convenience, becoming a true assistant for both life and work. I am absolutely convinced of this. Recently, Doubao has also made some attempts in this direction, sparking a lot of discussion, but I truly believe this is the right path.

As the glasses supply chain matures, in the future everyone will have a pair of smart glasses. Once you put them on, your cognitive abilities and work-life efficiency will be greatly improved. People won’t care so much about short-term issues like added weight or limited battery life—they’ll find all sorts of ways to accept or mitigate these, because iteration is ongoing, just like with the early days of smartphones. Early smartphones couldn’t run for a whole day—people either carried several devices or brought along power banks, but they absolutely couldn’t do without their phones. In the future, as large models become increasingly intelligent, people will similarly be unable to live without smart glasses. Phones may not be replaced immediately in the short term, but all the input and output interactions will shift to glasses. I am certain this day will come—probably sooner than we expect.

We haven’t tried to create a universal digital assistant for all work and life scenarios—instead, we want to empower sports scenarios with AI capabilities. For example, when working out together, the system can know your fitness level, understand your social circle, and proactively deliver value-added, proactive services. This could hugely improve the sports experience and make it less lonely. What we envision for the future is that sports glasses will become your personal coach, workout partner, and exercise buddy—that’s the future we’re looking forward to.

Wan: Next, I’d like both of you to look ahead: What other variables do you foresee in the smart glasses market? For example, not long ago Doubao came in and disrupted the field, and now companies like Tencent and others have directly banned Doubao; there are all sorts of restrictions on the mobile side as well. We will definitely keep seeing voices like these emerge one after another. Let’s start with President Yang for this challenge question: What’s your view on Doubao’s moves and the fierce pushback from the major players?

Yang: In my view, what Doubao is doing is exactly what every internet company should be focusing on, and I really appreciate this attempt. Exploration of new interaction models is irreversible—what might have seemed radical ten years ago, looking back ten years later, is absolutely a milestone. In fact, before Doubao collaborated with smartphones, we were already connecting large language models with smart glasses: products we launched last year already supported voice commands. For example, you could say, "Go to Taobao and buy the latest smart glasses," and it would simulate human gestures to bring up the page and complete the purchase. This kind of interaction eliminates all those layers of buttons and menus—just one sentence to get things done. I'm convinced it will become the core interaction method of the future.

As for tech giants imposing restrictions, honestly, that's normal. First, because there are still unresolved security loopholes: people around me have joked, "Transfer me 100 yuan," and the system might actually initiate a payment. We ran into this issue during our internal testing. Second, rights and responsibilities along the process—who collects the data, who stores it, who uses it—haven't been clearly defined. The boundaries between cloud service providers, content partners, and hardware manufacturers are still blurry. The backlash we're seeing today stems from the overwhelming number of unknown risks, much like when electric vehicles first hit the roads. But looking a decade ahead, this path is irreversible—that's my personal view.

Wu: Yes, I completely agree with Mr. Yang. This is really about exploring new ways of interaction—it's an iteration and a revolution in user interaction. All current solutions are just temporary; ultimately, it comes down to prioritizing user needs—does this interaction method truly improve user experience and efficiency? If it does, then it will definitely prove to be irreversible. The only questions are who will actually implement it, what the final solution looks like, and how benefits are distributed.

Let me add that, even so, you'll find that smart glasses are actually a more suitable entry point than smartphones. With a phone, you have to press a button, speak into it, and wait for something to appear—this isn't the most efficient approach. Ultimately, it’s likely that smart glasses will prove the most efficient solution.

Wan: Looking ahead, what do you see as the biggest challenges or obstacles in the development of smart glasses? What specific difficulties have they brought you? Facing these circumstances, how will you set your future strategies?

Yang: For smart glasses, starting from this year, no matter if you’re an industry insider or an ordinary user, there’s already a consensus—this will definitely be the next-generation device, and this is not in doubt in the long term. For startups, the biggest challenge is keeping the right pace: you can’t go too fast, but you also can’t go too slow. What does this mean? The supply chain still has its limitations, user education is not done yet, and there is still a technological gap between the product and the mass market—each of these has to be addressed one at a time. Looking ahead to the next ten years, every step—who to address first, who to tackle next—timing is crucial. Right now, what matters most is to deliver a solid experience for specific vertical users, make them feel it’s really useful, and at the same time accumulate long-term technical know-how, progressing hand-in-hand with leading international companies and the supply chain. It sounds easy to say, but when it comes to actually making decisions, the possibilities are so many—that’s what gives entrepreneurs real headaches.

Wu: Glasses are indeed the best hardware carrier for AI in the future. There are still some immature aspects of the current industry chain, so what we need to do is focus on the present—making use of the existing supply chain to address urgent needs in specific scenarios for specific users. We chose sports scenarios, because where the supply chain is lacking happens to be where we can offer better solutions.

Let me give an example: glasses right now weigh 40 or 50 grams, whereas regular glasses weigh only a dozen or twenty grams, so wearing them all day brings pressure to users. Is this functionality really an urgent need? If not, users will just try it for fun, but won’t really wear them all day long. We chose sports scenarios because, in sports, you naturally change glasses (cycling glasses, ski goggles), so the comparison isn’t with regular glasses, but with standard sports eyewear. Sports glasses are more wrapped around your head, the weight distribution is more reasonable, and you’re not wearing them all day—just for 3-6 hours of exercise, which is completely acceptable. This is about creating the right scenario to give users the best-possible experience and truly solve real pain points. Overall, we hope to leverage this wave of AI large models to genuinely reshape the sports experience and raise it to a whole new level.

Wan: We’ve also gone through that long iteration from Nokia’s keypad phones to feature phones—both of you are veterans of the mobile industry, and you understand what pacing means better than anyone. To wrap up, give me your prediction: how long before the ecosystem matures enough that consumers are truly willing to embrace smart glasses?

Wu: First, I think, at the current stage, smart glasses are probably where feature phones were. So secondly, between AI glasses and AR glasses, I still firmly believe AI glasses will become a mainstream consumer product much more quickly—they’ll really achieve mass adoption. My own estimate is that this could happen soon, maybe in two or even three years.

My assessment is based on three criteria: First, a single product model must be able to sell over ten million units; second, smart glasses must truly provide application services that users engage with frequently; third, the intelligent features must be powerful enough to enable truly proactive interactions. Based on these thresholds, I believe AI glasses could have the opportunity to become a mass-market consumer product within the next two to three years.

As for AR glasses, that timeline may be even longer. Having worked on AR glasses myself, I think the entire industry chain and app ecosystem still have quite a way to go before reaching maturity. If AR glasses are to truly become a mainstream consumer product, it might take three to five years, or perhaps even longer. That’s my assessment.

Yang: I completely agree with Mr. Wu’s timeline—AI glasses will enter the mass market relatively quickly within two to three years; but true AI+AR glasses that rival the next generation of smartphones may take three to five years, or even longer, before they see widespread adoption. We’re currently in the phase of serious uphill effort.

Wan: Today’s discussion has made me even more convinced: AI glasses could bring about brand-new opportunities. I hope our audience also starts thinking—this consumer product we’ve always been familiar with, once enhanced with vision, AI, and various functionalities, might just become an essential assistant for you to interact with the world in the future.