Wei Zhe, former CEO of Alibaba and later founder of Jiayu Capital, raised a question to Zhu Mingming, founder of AR company Rokid, at a gathering: You are now at the forefront of the industry, but can you think of a core issue? What does it mean if a top company only earns a few hundred million yuan in revenue per year?
There are two reasons why Wei Zhe asked this question: first, this is already a market that is not large enough; Secondly, it indicates that the industry has not found a real explosive point. This interpretation is fatal for any track, and it is no exception for AR.
Today, driven by the popularity of Apple's release of Vision Pro (AVP), the popularity of AR has risen again, and the number of developers in the Rokid developer community has increased from over a thousand to over three thousand. However, there is still a long way to go before the market matures and scales up - for Rokid, as well as for Apple's AR business.
Recently, Rokid launched the new AR glasses Rokid AR Lite. During an interview with First Financial reporters, Zhu Mingming talked about revenue growth, his views on AVP, his attitude towards AI big models, and the future trends of the AR industry. In Zhu Mingming's view, the current AR industry is very similar to the first three years of the birth of the iPhone. All practitioners are exploring, but who will become the rising Apple? Who will become the declining Nokia? This is where no one can answer but is also anxious.
Apple Vision Pro once again heats up the AR industry
The AR track inevitably talks about Apple's newly launched Vision Pro.
The period from 2014 to 2016 was a period of low-cost popularization of VR concepts, but due to immature technology and lack of ecological support, related concepts quickly stalled. Nowadays, with the explosion of Apple's concept, AR is once again popular, and technology giants have also listed XR as a carrier for space computing.
The industry had high expectations for AVP sales performance, but the initial release did not meet expectations. In Zhu Mingming's view, the release of AVP can be considered very successful, as it is a strategic adjustment of Apple's product definition over the years. During the Steve Jobs era, Apple made all decisions before users received the product - what to use, what not to use, how to use it, and what not to use it. But after Tim Cook, AVP showed a clear "heap functionality" style, with the market providing feedback on what is needed and what is not. In Zhu Mingming's view, this may be the most suitable thing for the current Apple supply chain to do.
Zhu Mingming told reporters that the presentation and release of AVP's products this time adopted a very stable approach. "As an entrepreneur, I feel that Apple is very intelligent. This is also the essential difference between Apple today and the past. With a very clear understanding of itself, it has changed the way it defines products."
Technically speaking, AR glasses are divided into VST (Video see through) route and OST (Optical see through) route. The former, such as AVP, mainly provides an immersive sensory experience by inputting external visual information through the camera, but has the disadvantages of heavy equipment and isolation from the outside world. The latter, such as Rokid, allows people to see the real world while overlaying and constructing virtual information, which has the advantage of being more lightweight while also facing issues such as limited computing power.
Huatai Securities believes that although OST has the characteristics of being lightweight and energy-saving, there are few manufacturers that can truly introduce OST solutions, mainly limited by the complex optical path design of optical perspective, high production difficulty, and high cost of optical components.
As a startup, Rokid chose the OST route. Zhu Mingming told reporters that on the one hand, the company itself recognizes this route. Secondly, objectively speaking, VST is a highly comprehensive route that must be completed by large companies investing heavily in resources and large-scale team operations, including chips, sensors, material optics, etc. In today's industry, apart from Apple, even Meta has not been able to perfectly showcase related products.
Public information shows that AVP is equipped with two chips, M2 and R1. M2 provides strong computing power, while R1 reduces device latency and elevates computing power to the computer level, making spatial computing possible. The R1 chip is designed to handle real-time sensor processing tasks, responsible for processing data from 12 cameras, 5 sensors, and 6 microphones, solving latency issues, and ensuring that new images can be transmitted to the display screen in an extremely short time (12 milliseconds).
In addition, from the perspective of market influence, Zhu Mingming told reporters that Rokid does not have the same appeal as Apple, making it difficult for consumers to spend tens of thousands of yuan to buy a technology product that is not perfect but looks good in some aspects. But on the B-end, as long as it can solve the core pain points of specific scenarios, it is acceptable, which is also Rokid's current main revenue direction.
Although AVP has been somewhat lackluster, track companies including Rokid still have high expectations for Apple. Zhu Mingming told reporters that he hopes Apple can launch lighter products as much as possible, benefiting the entire industry chain. Although consumers have many more options in the mobile phone field, and even to some extent, they have achieved a certain degree of disenchantment from Apple, in the AR field, Zhu Mingming believes that it is not yet possible to leave Apple. The most core point is that everyone seems to forget that beyond market value, Apple is currently a huge semiconductor company on a global scale, and almost every Apple phone and computer has an Apple chip. Although the OST route does not rely on computing power support like VST, the industry appeal triggered by the scale of Apple's AR products is currently the hope of the early track.
Industry Anxiety AR "iPhone Moments"
For many years, Zhu Mingming did not explicitly respond to financial related data, and product development and technological exploration were put at the forefront of his demands. But in recent years, under the increasingly strict capital environment, as a startup company, Rokid has fallen from technology flow to pragmatism.
Zhu Mingming stated that Rokid's sales increased by 150% year-on-year in the past quarter, and this year's target sales will reach 300000 to 500000 units. Compared to the consumer electronics industry, this sales range is sufficient to match the sales of high-end models. If AR can implement this data, it can achieve a certain degree of breaking through the circle.
Rokid, founded in 2014, started generating revenue in 2021 and saw an eight fold increase in business revenue in three years to 2023. Starting from 2022, the company will adjust its strategy from prioritizing scale to prioritizing profit. The current Rokid CEO meeting has undergone significant changes compared to before 2019. In the past, it was about technology and products. Now, Zhu Mingming will ask the team, what benefits will a certain technology investment bring to the company within 12 months?
As for the current hot large-scale AI technology, Rokid is not in a hurry to invest. Zhu Mingming told reporters that first of all, Rokid has benefited from AI's technological development, but in his opinion, the current AI track is still in the foam period. Taking the mobile phone and computer industry as an example, this field requires new hot stimulation and finding exciting "small pills".
"But at present, AR technology is fresh enough and has not even been fully implemented, so there is no need to add a technology that may not generate rigid demand." Zhu Mingming told reporters that Rokid's current focus is on how to make users try to accept AR products first. Assuming that after a period of time, consumers truly accept spatial computing and AR interaction methods, this technology makes users accustomed to it. At that time, perhaps it can be said to the public that "AR is the best carrier of AI".
From the perspective of the entire industry, the most mature operating system for Rokid is currently the operating system, while the least mature or still in the cultivation stage is optics. Under the limitations of optical technology, wearability has also become a factor affecting landing. Both VST and OST manufacturers are trying to make their devices lightweight, but the challenge remains significant. Zhu Mingming told reporters that the exploration of both technological routes in terms of lightness and thinness takes a period of five to ten years to complete. However, it can be affirmed that spatial computing is definitely the next generation of computing platforms after desktop and mobile computing. As a startup company, the primary task is to survive.
In addition, Zhu Mingming stated that although the ultimate large-scale popularization of a device mostly occurs in gaming and social scenes, the AR industry is not yet ready because the fragmentation problem of AR devices has not been solved. The so-called "fragmentation" refers to providing immediate feedback upon receiving a message from the other party, just like chatting on QQ at any time. But nowadays, whether it's VR or AR, the device needs to be removed first and then worn, and the possibility of being fully online 24 hours a day is very low.
Zhu Mingming said that Rokid naturally hopes that social functions will grow within its own ecosystem, but it needs to be taken slowly and restrained. "Otherwise, the company's small amount of money will burn out in half a month."
Zhu Mingming likened the current AR industry to the first three years of the emergence of the iPhone. Everyone was defining smartphones, and the industry saw products such as Nokia flip phones, BlackBerry phones, and Motorola handwriting machines. The same goes for the current industry, and practitioners including Apple cannot find a clear answer. Zhu Mingming analyzed that there will be three situations in the future: large companies educate the market very well, and the two technological routes do not directly compete with each other, both have survival space; The second type is that the market is well-educated, but there is also full competition among products, and start-up companies are facing fierce competition from large companies; The third scenario is the worst-case scenario, where even large companies fail to educate the market and the entire industry is destroyed.
If there is anything else that keeps Zhu Mingming awake now, it is the question of "Have I found the answer?" Wei Zhe's question is difficult to answer, and the current situation is also difficult to break through. "No entrepreneur can answer future questions. All we can do is try to stand at the door when that door opens, even though I don't know which one is right now." Zhu Mingming said.