VITURE, which once ranked first in the North American AR glasses market, said, "AI and AR will eventually merge into a new generation of multimodal devices."

Text | Wang Fangyu

Editor | Su Jianxun

In the current booming era of AI glasses, the brands in the previous hot - track of AR glasses are facing a common choice: Should they enter the AI glasses market?

Some brands choose to take the initiative. For example, the first - generation AI glasses of Thunderbird Innovation and Rokid were released last year. Some brands are still in the stage of waiting and preparation.

VITURE belongs to the latter. In the view of Jiang Gonglue, the founder of VITURE, "The threshold and difficulty of making good AI glasses are very high. One must think carefully and be cautious."

Jiang Gonglue graduated from the School of Design at Harvard University, majoring in human - computer interaction. He has engaged in human - computer interaction research at Microsoft Research Asia, MIT Media Lab, and Google X, and participated in the project exploration of Google Glass.

Intelligence Emergence exclusively learned that by the end of 2024, VITURE completed a financing of $50 million. The investors include Wang Huiwen, the co - founder of Meituan, and Singapore Telecommunications Group.

Jiang Gonglue, the founder of VITURE. Source: Authorized by the enterprise

Since its launch in October 2023, Ray - Ban Meta has successfully sold more than 2 million pairs, heating up the AI glasses market. In contrast, despite years of development, the global sales volume of AR glasses was only 500,000 units in 2024. The former has a larger market space and more imagination.

As the founder of an AR glasses brand, Jiang Gonglue is highly optimistic about AI glasses and has conducted continuous observation and research.

Jiang Gonglue told Intelligence Emergence directly, "AI glasses on the market, including Ray - Ban Meta, are not really AI glasses yet. Users buy Ray - Ban Meta for the Ray - Ban brand and its photo - taking function, which has little to do with AI. The real AI glasses have not been proven by the market yet."

In his view, the core and "soul" of AI glasses should be AI capabilities, but the AI functions of current products on the market cannot satisfy users. This is a gap and opportunity to be filled and improved.

"Today's AI is just a Chatbot that can only chat with users. The ideal AI in users' minds should be an assistant that can help users send text messages, order takeaways, and complete almost all the work that can be done in today's mobile Internet and digital world. This is the real value that AI glasses bring to users."

VITURE's AR glasses. Source: Authorized by the enterprise

The reason why VITURE didn't enter the market earlier is that Jiang Gonglue saw the complexity and difficulty of making a good AI glasses product.

He said that practitioners tend to underestimate the difficulty of making glasses. It's not just about assembling the supply chain. There are many details in smart glasses that users will complain about. If these details are not solved well, it's difficult to create a best - selling product.

Compared with hardware polishing, product definition is a more critical proposition.

Jiang Gonglue believes that to define a good AI glasses product, it is necessary to clarify the target user group and wearing scenarios without any ambiguity. Otherwise, a huge mistake will be made.

People's upper - limit expectation for AI glasses is "to become the next smart terminal to replace mobile phones", but Jiang Gonglue holds a negative attitude. He believes that glasses will not completely replace mobile phones, just as mobile phones did not replace computers. The end - result will be an organic integration.

The relationship between AI glasses and AR glasses is the same. He predicts that when AR glasses become lighter and AI glasses have more functions, there will be an intersection and integration between the two, just like two slopes leading to the top of the mountain. This intersection point will be the multi - modal interaction device with display, camera, and microphone that countless users dream of, with a perfect experience.

"AI glasses and AR glasses will eventually integrate in the future." Jiang Gonglue told Intelligence Emergence.

The following is the edited transcript of the dialogue between Intelligence Emergence and Jiang Gonglue, the founder of VITURE:

"Ray - Ban Meta is not AI - enough. AI capabilities can't just be a Chatbox"

Intelligence Emergence: You come from the AR glasses industry. How do you view the opportunities in the AI glasses market?

Jiang Gonglue: AI has become as ubiquitous as water, electricity, and air. So I think AI glasses will definitely become a necessity for people. As a carrier of AI, we believe that AI glasses can form a large - scale market, which may be larger than that of AR glasses in the future.

AI glasses and AR glasses will eventually integrate in the future. When AR glasses become lighter and AI glasses have more functions, there will always be an intersection point. This intersection point is the ultimate multi - modal interaction device with display, camera, and speaker that everyone dreams of, which can also meet the requirements of weight and battery life. We believe this can definitely be achieved in the future.

Intelligence Emergence: How do you evaluate the AI glasses products on the market, including Ray - Ban Meta?

Jiang Gonglue: Actually, the real AI glasses have not been proven by the market. People don't buy Ray - Ban Meta glasses for their AI functions. Meta has two biggest labels. One is the Ray - Ban brand. If you regard it as an SKU of Ray - Ban sunglasses, it's not hard to understand the rationality of its sales volume. The other label is photo - taking. It's a head - mounted sports camera similar to DJI and Insta 360. This is a proven market. Meanwhile, the functions of photo - taking and listening to music are perfectly combined with the scenarios of sunglasses.

Intelligence Emergence: You said that Ray - Ban Meta is not AI - enough. What will the real AI glasses be like?

Jiang Gonglue: The real AI glasses need to have three key points:

One is that it needs to be infinitely close to normal glasses, so that people are willing to wear them all day long, and there should be multiple SKUs to meet personalized needs.

Second, the sales channels of AI glasses are very important. Now, peers are all cooperating with optical stores. I think this is the right idea. Selling AI glasses can't rely only on online channels.

Third and most importantly, it's the AI capabilities. The ideal AI in people's minds is an assistant, a real AI Agent. But today's AI is just a Chatbot that can only chat with you and can't help you complete any work.

For example, if you say "Send a message to my mom saying that I'll go home late for dinner tonight", current AI glasses can't do it. The real AI should be able to help users complete almost all the work that can be done in today's mobile Internet and digital world. This is the real value that AI glasses bring to users.

At the same time, AI glasses can continuously understand users' needs. It has your memory and can understand you like an assistant who has followed you for three or five years.

Ray - Ban Meta. Source: Visual China

Intelligence Emergence: Do you think AI glasses can replace mobile phones in the future?

Jiang Gonglue: Many people are saying that AI or AR glasses will replace mobile phones in the future. I've always been asked this question, and my answer has always been "No".

The organs through which humans interact with the outside world are only eyes, ears, mouth, and hands. The future form must be a combination of mobile phones and glasses. Glasses will be responsible for the mouth, eyes, and ears, while mobile phones will be responsible for the hands. It's not about one replacing the other. This is a natural form, not something that can be forced to happen.

However, when the experience of AI glasses matures, AI will break down our needs into commands that computers and mobile phones can understand, which will greatly reduce the frequency of using hands. So I believe that AI and AR glasses will replace most of the functions of mobile phones, and people's interaction interface will gradually shift from mobile phones to glasses.

Intelligence Emergence: How is VITURE's AR product combined with AI at present?

Jiang Gonglue: VITURE has also made a lot of explorations in AI very early. The 2D - to - 3D conversion is the most popular function on VITURE glasses at present. It can automatically convert pictures and streaming videos into 3D in real - time, which enriches the content of AR glasses. We were recently invited to NVIDIA's GTC conference to showcase this function. Its realization benefits from the progress and combination of AI capabilities and AI computing power.

In addition, we have a game guide assistant called Wizard, which is also based on a large - scale AI model. When users are playing games, they can directly ask the Wizard assistant how to defeat the enemy at any time. We use game guides as a database and combine RAG (Retrieval - Augmented Generation) technology to achieve the experience of chatting while playing games.

"The threshold and difficulty of AI glasses are very high. Hardware, software, and AI capabilities need to be coupled"

Intelligence Emergence: Is there any difficulty in entering the AI glasses market from the AR glasses field?

Jiang Gonglue: I think people often underestimate the difficulty of making glasses. It's not just about assembling the supply chain. There are many details that users will complain about. If these details are not solved well, it's difficult to create a best - selling product.

Theoretically, competing with Ray - Ban Meta in making AI photo - taking glasses requires a large amount of resources. Some manufacturers may have a gap between what they want to do and the resources they need, resulting in not achieving the expected results.

Just for photo - taking, the photo - imaging teams of leading mobile phone manufacturers have thousands of people. This ability can be reused for glasses. So sufficient resource investment is needed to do this well. But I also believe that through some iterative updates in the future, the experience will get better and better.

Intelligence Emergence: What about other types of AI glasses? Can you elaborate on the details?

Jiang Gonglue: Glasses are different from mobile phones, tablets, or AI toys. There are many dimensions to consider when making glasses. The space of glasses is very small. It's easy to run into difficulties when trying to fit so many structural and electronic components. Otherwise, the glasses will be very bulky, and no one will want to buy them. Glasses are also a movable, bendable, and elastic electronic product, which requires high reliability.

Finally, the most critical thing is how to balance performance, power consumption, and battery life. It's not like other devices where you can just install a large battery on a board. In the future, there may be 7 - 8 batteries on AI glasses. Batteries will be installed wherever there is space to ensure the ultimate product form.

In addition, to make good AI glasses, one not only needs to have hardware capabilities but also strong software capabilities and AI capabilities. The three need to be coupled together to succeed. For example, to achieve different levels of AI capabilities, the requirements for the SOC capabilities of glasses are different, and the corresponding power consumption during use and standby is also different, which puts forward different requirements for the glasses' batteries. The hardware, software, and AI capabilities are integrated end - to - end.

VITURE's AR glasses. Source: Authorized by the enterprise

Intelligence Emergence: How does VITURE consider the usage scenarios of AI glasses?

Jiang Gonglue: I think the AI assistant mentioned above is the most important scenario, which is more of a productivity tool. Different from previous audio glasses, which only have the functions of making calls and listening to music and do not have AI capabilities.

In terms of the target population, there is a classic four - quadrant model: divided into the general population and niche populations, and scenario - based wearing and all - day wearing. Meta is more for niche users' scenario - based wearing, which is similar to AR glasses. This is a good starting point. The next step will be in two directions: all - day wearing for niche populations and scenario - based wearing for the general population.

Regarding the difficulty level of these two quadrants, we think the all - day scenario for niche populations is better. Usually, a product penetrates the market from a certain niche population from scratch.

Intelligence Emergence: What daily interaction scenario requirements do you have for AI glasses now?

Jiang Gonglue: I have quite a lot of interaction requirements. For example, I ask it to recommend a song for me, or I hope to have some light music to help me concentrate at work. Another example is asking it to read or summarize a WeChat official account article. And I also want it to help me order a cup of coffee.

In summary, in essence, it is a more efficient way of interacting to obtain information and use functions on mobile phones.

Intelligence Emergence: If VITURE wants to make AI glasses, do you have any different ideas?

Jiang Gonglue: With VITURE's approach, we will first select the target population and market, then understand the channel pattern and technological trends, and finally determine the functions and Product - Market Fit (PMF). Instead of looking at what kind of camera, chip, or display solution to use at the beginning.

Even when using the same BB (Birdbath) solution, VITURE has made its AR glasses differentiated by developing a variety of popular accessories and software adaptations. The same goes for AI glasses. The body structure may be similar, but the software ecosystem and functions are the differentiating points. Whoever can organically combine the AI agent with the glasses well can achieve the best PMF.

Another aspect is the appearance design. Smart glasses are a product that pursues the ultimate balance between functionality and wearability. Glasses have a fashion attribute. It's difficult to design them to be fashionable, and it's even more difficult to lead the trend. It's not something that can be solved by just finding a design agency. This requires the team to have a DNA of design and aesthetics, and a humanistic, artistic, and even spiritual understanding and belief in integrating technology into life, so as to integrate this understanding into the product and convey it to users to resonate with them.

"There is no leading brand in the AR glasses market. The core is to explore new growth opportunities"

Intelligence Emergence: You have been in the AR industry for many years