HarmonyOS has achieved half of its success.

There are two interesting discussion trends surrounding the progress of HarmonyOS. A few weeks ago, Huawei announced that the number of HarmonyOS 5 terminals exceeded ten million, and its share in China surpassed that of iOS, making it the so - called second - largest operating system.

Scale and quantity are crucial. Without them, it's impossible to attract more developers to port their applications to the HarmonyOS ecosystem, and Huawei won't be able to generate high - value internet revenue.

Huawei adopts a similar strategy to Apple. Externally, it has a unique and attractive hardware system, such as triple - folding phones and foldable PCs. Internally, it has a closed development ecosystem, with the core being a proprietary and closed operating system and chip system.

The most obvious weakness is the chip manufacturing process, while its strengths lie in hardware design and the market. The brand - new HarmonyOS ecosystem was once a weak point too. Huawei's terminal hardware, including phones, cars, and PCs, has to support two ecosystems: one is the gradually improving but long - lagging chip manufacturing process, and the other is the HarmonyOS ecosystem starting from scratch.

HarmonyOS: Old Wine in a New Bottle

Supporting either end requires a sufficient market scale. It provides opportunities for the improving chip manufacturing process to try and error and offers sufficient attractiveness to HarmonyOS ecosystem developers. Not long ago, Ren Zhengfei, in a rare interview with a media outlet, mentioned this topic and innovatively proposed the concept of "three compensations": "Non - Moore compensates for Moore, mathematics compensates for physics, and group computing compensates for single - chip."

There is another compensation that Ren Zhengfei didn't mention, which is using the market to compensate for technology. By offering various unique and attractive hardware products to attract users and gain a large enough market, non - Moore and mathematics can come into play.

The HarmonyOS terminal ecosystem of the Huawei ecosystem acts like a locomotive, pulling the chip manufacturing capabilities of the Huawei ecosystem. It's not easy to integrate hardware with a lagging manufacturing process into the current mainstream mobile internet ecosystem. Recently, data disclosed by Huawei shows that thousands of applications have been adapted and developed for the HarmonyOS version.

The increase in the scale and share of the HarmonyOS ecosystem is of great significance to Huawei. It allows Huawei's hardware terminals to maintain a "fertile ground" - like existence, continuously bringing substantial internet profits to Huawei. Apple and other Android manufacturers currently highly rely on this business model.

The interesting part is here. An operating system based on AI technology applications, considering several strategic interests, rapidly expands its scale and is backward - compatible with the advertising and distribution business models of the mobile internet era. Does it still have the speed and trend advantage for forward evolution?

What are the biggest technological and industrial trends? Li Kaifu, Jian Lifeng and other former Googlers have the most intuitive feelings. Large - language models turn web pages into answers. Without links and web pages, the foundation for web page and advertising bid - ranking becomes meaningless.

This directly impacts the business models of search engine companies like Google and Baidu. That's why these two business giants, which led the development of computing clusters and large - model frameworks earlier, will never be able to make breakthroughs in this path.

The problem is that the business models of Apple, Huawei, and Android phones currently rely heavily on Google and Baidu. They are also typical bid - ranking business containers. That's why phone companies are not very determined about AI, with Apple being the most typical example.

It's not just a technical issue. More importantly, it naturally conflicts with the existing business models of these giants, and no one has the real courage for self - revolution.

Originally, Huawei had a huge advantage. Due to industrial chain bans, its market share dropped sharply, and it was unexpectedly freed from this golden handcuff. As the share and scale of the HarmonyOS ecosystem rapidly increase, Huawei phones have once again become a valuable source of monetization for Huawei.

However, at the same time, it has also picked up a business burden that is difficult to abandon easily.

Against this background, the HarmonyOS ecosystem will pay more attention to large - scale killer application manufacturers that can immediately bring advertising and distribution revenues, rather than focusing on cultivating and discovering new application manufacturers based on AI, let alone creating new AI applications on its own.

Just like what Sam Altman and Jony Ive did, Yu Chengdong doesn't like such strategic experiments. As the locomotive, he has a heavy KPI to fulfill.

Tim Cook, Lei Jun, Shen Wei, and Chen Mingyong also won't really like such new things, regardless of how they describe black technology and new species. In essence, they hate new technologies that bring about new changes and those that will disrupt the existing mobile phone landscape.

In the subconscious of successful people, they hope the technological landscape will stay as it is.

The only and best possibility is that Huawei and major Android manufacturers may choose to use a new brand to conduct small - scale and quiet experiments to explore what the future AI phone will be like and actively face the challenges from people like Jony Ive. Coincidentally, several highly influential AI hardware teams have encountered various difficulties in their entrepreneurship.

This article is from the WeChat official account “Intelligent Things Technology Review” (ID: IntellegentThings). Author: Wen/Zhiwu. Republished by 36Kr with permission.