Source: Wall Street See
Author: Zhou Yuan

The allure of AI technology is unstoppable.

The upcoming Apple iPhone 16 series, which is set to be released in September, is expected to have the biggest highlight of a software system upgrade based on AI technology, given its relatively lagging hardware compared to the Android camp. At the same time, Google, the developer of Android, the archenemy of iOS, also unveiled some AI features of its Pixel 9 series on August 13th.

In the era of "traditional smartphones," Google, the developer of the Android system that developed iOS's biggest competitor, had almost no presence and few people noticed that Google also had a smartphone. In the era of AI, however, Google may have the opportunity to change this embarrassing market perception.

The market has reached a consensus that the iPhone 16 series will be Apple's first AI phone. Do not underestimate this. After all, the Samsung Galaxy S24 also had limited hardware upgrades, but its sales skyrocketed thanks to the AI phone label.

The biggest highlight of the iPhone 16 series is AI.

Although there was nothing new about Apple's Apple Intelligent strategy, which used AI technology to rebuild the thinking of multiple operating systems, including iOS 18, at WWDC 2024, it actually replicated the first and second layers of the 'AI Four-layer Architecture' technology framework of Chinese brand Honor.

Due to its brand, technological innovation, and product definition appeal in consumer electronics, Apple's technology route, which is favored by Apple, is also unrivaled in the global industry, and its final market acceptance is similarly unparalleled. This can be seen from Apple's stock price, which fell 1.91% (total market value of 296.05 billion US dollars) on the day it released its AI strategy at WWDC 2024 (June 10), but continued to rise by a cumulative 16.89% (total market value of 347.07 billion US dollars, second only to Microsoft's 3.475 billion US dollars, making it the second-largest technology company in the world) from the next day (June 11) to July 5.

So, what AI functions will the iPhone 16 series have? First, let's talk about the very limited hardware upgrades of this generation.

In addition to the exclusive news from Wall Street on the battery case of the Apple iPhone 16 series being changed from aluminum alloy to small steel shell (with battery capacity remaining unchanged at only about 4600mAh/40W), this generation of Apple's flagship has also upgraded its imaging system to an upgraded 48 million-pixel ultra-wide-angle lens, and has also upgraded the 3x upright long-focus lens to a 5x periscope long-focus lens. The screen uses Samsung's new M14 panel, and the peak brightness in some areas has been increased to 2000nit. Starting memory is also standard at 8GB.

This kind of hardware upgrade is so small that it can be easily surpassed by China's numerous Android counterparts. Therefore, AI has become the market focus of the iPhone 16.

Apple's decision to upgrade the memory standard of this generation of flagship to 8GB is also based on AI considerations. Because Apple Intelligent requires devices to have internal memory starting at 8GB.

The A18 chip of the iPhone 16 standard version and Pro version and the A18 Pro chip of the Pro Max version are all made using TSMC's 3nm process technology. The only difference is the difference in advanced technology characteristics, such as AI computing power and ray tracing performance. For the end-user experience, these differences are not clearly perceptible. Therefore, pricing is "difficult": the hardware upgrade is limited, and the application of software is not clearly perceptible.

However, from Apple's "innovative" AI perspective at WWDC 2024, features like generating abstracts from speech-to-text on iOS, Siri calling ChatGPT for free (associating with GPT-4o), answering questions more naturally, understanding ambiguous expressions, being able to handle multi-turn dialogue, AI photo editing, AI video, and photo creation are not amazing.

According to John Giannandrea, Apple's Senior Vice President of Machine Learning and AI Strategy, and Craig Federighi, Apple's Senior Vice President of Software Engineering, the large language model running on iPhone is about 3 billion, which is significantly lower than the domestic counterparts' starting point of 7 billion for the end-side large model. The gap is evident.

Of course, it's difficult to say what AI features the iPhone 16 series will have and whether they will be "amazing."

Google is also promoting AI phone technology.

Looking at Google again, AI applications in the upcoming Pixel 9 series are not built into the Android system, but come from Google AI (corresponding to Apple Intelligent?).

According to leaked information, the AI features of the Pixel 9 series mainly include three categories: Add Me, Screenshots and Studio, corresponding to AI photo editing, AI screenshot information search, and AI creation studio. In terms of core features, Google believes that the role of end-side AI should be centered around C-end users' daily scene applications, which may not be concentrated in the "productivity tools" level as claimed by domestic vendors.

The AI application of Screenshots, in fact, also combines Google's founding function, which is search. From the leaked information, Screenshots search is mainly for searching for relevant information about historical screenshots, making it easier for users to quickly find target images from thousands (or tens of thousands) of screenshots on the end-side based on fuzzy memories.

In addition, like Huawei, Google and Apple are also promoting the integration of AI technology and Android system. Huawei calls it "restructuring the underlying operating system with AI technology". If Google integrates AI technology into the Android system, what is the difference between the Pixel 9 series and the previous Google inventory smartphones in terms of AI application?

Industry speculation suggests that if the Pixel 9 series is used, AI functions can be used on the client side, while the inventory Google phones can only call AI on the cloud side due to chip performance and memory capacity limitations, and the speed may be slower. Once disconnected from the Internet, AI applications cannot be called.

If client-side AI applications are attractive enough and can frequently meet the urgent needs of users' scenarios, then a new wave of phone upgrades may occur.

On this point, Google is very similar to Huawei's platform-level AI function called "Any Door" launched in January this year: both focus on user daily applications, rather than "office productivity tools".

For example, some AI applications of Google Pixel 8 series: the quick information screening function allows users to call Gemini by just moving their mouths and inquire about information in Google's related applications, and then directly call the corresponding APP services.

Therefore, Apple and Google, the developers of iOS and Android systems, are looking forward to a new technology application exam for AI smartphones, which is already on the horizon. Whether their respective AI smartphones can inject new vitality into the stagnating inventory market due to the lack of innovation in new technologies, will soon have an answer.

From this perspective, Apple and Google are working together to usher in a new technological journey for smartphones in the AI era.

Editor/Jeffy