Summary of Key Points
The AI glasses industry is currently in a state of intense activity (for example, dozens of manufacturers filled the exhibition halls at the WAIC conference), and China's supply chain has become so robust that it’s almost like building glasses by simply assembling components (China accounts for 80% of global manufacturing shares, with upstream chip companies experiencing profits that have increased by several dozen times). However, the industry is stuck on three critical issues: hardware cannot simultaneously achieve being lightweight, durable, and powerful; privacy concerns regarding cameras deter users from wearing them; and most features become boring after a few uses, leaving users without a reason to wear them long-term. More importantly, no one has been able to clearly define what AI glasses are—Are they sunglasses? Cameras? Or a type of phone that doesn’t require hands to use? Large companies are experimenting with different approaches, but none have yet convinced the market.
1. China’s Supply Chain: The “Super Factory” for AI Glasses
You might be surprised to learn that you can now buy an AI glasses set with a camera for just 300 yuan in Shenzhen’s Huaqiangbei area. This isn’t due to some miraculous technology; it’s because China’s supply chain has already prepared all the necessary components, allowing for easy assembly.
- Speed is astonishing: Owners of ODM (Original Design Manufacturing) factories in Shenzhen say that customers can get samples the same afternoon they submit their modification requests, and the shortest time from project initiation to mass production is one and a half months. Why? Because most of the components are ready-made “off-the-shelf” parts; all that’s needed is to select the desired functions and reduce the weight.
- The supply chain is so well-connected that it’s almost effortless: 80% of global AI glasses manufacturing takes place in China, with components coming from Dongguan for lenses, Jiangxi for mirrors, and Shenzhen for assembly. Suppliers are within a two-hour drive of each other. Meta’s Ray-Ban glasses are manufactured by Goertek, which relies on Sunny Optical for optics, and the chips used come from Hengxuan Technology—these companies together form a network that ensures issues can be resolved immediately.
- Upstream companies are already reaping huge profits: Chips and optics account for the majority of costs (70%), and upstream companies are benefiting first. For instance, Baiwei Storage’s profits increased by 32 times in the first half of the year, and Shannon Xinchuang’s by 21 times, due to the surge in demand for AI glasses.
However, this also brings problems: as the supply chain matures, products become increasingly similar—any advantages in weight or features can be quickly matched by competitors within six months, leaving users unsure which brand to choose.
2. Three Major Barriers: The “Deadlock” of AI Glasses
The industry appears lively, but it’s actually constrained by three major hurdles that are interlocked:
1. The hardware impossible trinity: You can only achieve one of the following three goals—lightweight, durability, or power. Want lightweight glasses? Li Weike reduced the weight to 26 grams by removing the display; Moonix cut it down to 14.9 grams, even sacrificing a camera. Need long battery life? Glasses with displays last at most 5 hours, while screenless models can last up to 8 hours (which is why screenless models saw a 463% increase in sales in 2025). Want high performance? Increased chip power consumes more battery life. These are physical limitations: battery energy density only increases by a few percentage points annually, while AI computing power demand doubles yearly, making a breakthrough difficult in the short term.
2. Privacy concerns: Cameras are a double-edged sword. The core value of AI glasses lies in their cameras (for translating menus, navigating, etc.), but they also act like mobile surveillance devices. Harvard students conducted an experiment using Ray-Ban glasses and found that taking photos of strangers could provide their address and phone number within minutes. Meta added a flash for photography, but a 9.9 yuan sticker can easily cover the camera. This issue cannot be solved by technology alone; societal rules are needed to determine whether and where cameras can be used in public places, and how data will be managed. Without clear regulations, users are hesitant to wear such glasses, limiting their practical use.
3. User retention: AI glasses offer over 200 functions, but less than 6% of users use them regularly, with a high return rate on TikTok (50%). Why? Functions like translation are only useful when traveling abroad, and photography features become boring after a week. Voice assistants for checking the weather are no better than using a phone. Users find the products “cool” at the time of purchase but realize they’re not very practical. The industry focuses on what it can do, while users wonder why they need it—this gap remains unbridged.
These three barriers combine to create a vicious cycle: poor hardware leads to users avoiding them; privacy concerns prevent further adoption; and lack of demand stops brands from investing in improvements to the hardware.
3. The Lack of a “iPhone Moment”: Who Can Define AI Glasses?
Before the iPhone, all phones had buttons, but Steve Jobs redefined the concept with its touchscreen and apps. What AI glasses need now is a similar defining moment. ByteDance’s first-generation product was canceled because it looked too similar to Xiaomi and Leino products, leaving 50 million yuan in research and development and 100,000 units unsold. Large companies are each exploring their own directions: Alibaba calls them “intelligent glasses” (proactive services), Li Weike calls them a “second brain” (for memory assistance), ByteDance’s second generation aims for extreme lightweightness (under 20 grams), and Meta says AI glasses are simply glasses with added intelligence (sold for $299, resembling sunglasses). The winner will be the one who can clearly define the category. In 2026, global sales of AI glasses are expected to reach 16 million units, with Meta holding a 60-70% market share due to its clear definition. Apple plans to release its own AI glasses in 2027, and Google and Samsung are also making moves. The company that can succinctly explain why users must buy these glasses will hold the key to this billion-dollar market.
Conclusion: Behind the Hype, There’s Silence
The WAIC conference was crowded, but the real breakthrough has yet to happen. There are AI glasses on the market, but there’s still no clear “category” for them. It was similar with MP3 players before the iPod; search engines didn’t become popular until Google emerged. All manufacturers are still exploring in the dark. The one that can find a compelling reason for users to buy AI glasses will become the next Steve Jobs. Until then, all this excitement is just trial and error.
---
Please note that this translation aims to maintain the original structure and tone of the Chinese analysis while adapting the language to fit English financial journalism.