Solving the Screen Island Problem with AI Connectivity

Connect Everything with AI

Connecting Everything with AI

In 1995, Mamoru Oshii adapted Masamune Shirow’s sci-fi manga “Ghost in the Shell” into a theatrical release, where the protagonist, Motoko Kusanagi, is one of the first humans to undergo full-body cyborg transformation. In the movie poster, Kusanagi is depicted nude, connected to mechanical devices, except for her brain and part of her spinal cord. This was a fantasy and metaphor by sci-fi authors from two or three decades ago about human-device coexistence.

Time has passed, and nearly thirty years after the release of “Ghost in the Shell”, despite our lives still being far from the cyber world, how to better connect with devices has become an unavoidable issue.

Screen Overload Problem

Deloitte’s 2023 “Internet Consumer Survey Report” states that the average American household owns 13 types and 21 devices, a 20% decrease compared to the home office period in 2021, with more people facing a “device ceiling.” People are beginning to realize that owning too many devices is not beneficial; the key is to better connect these devices.

This is not an easy task.

Almost at the same time Masamune Shirow began serializing “Ghost in the Shell”, Microsoft launched the Microsoft Office suite on Apple’s Macintosh computer — now the largest software and hardware company, also a pioneer in exploring better connections between devices and humans. During the PC internet development phase, a person had only one computer, which was an effective method for handling certain tasks with a single device.

However, in the mobile internet era, old methods no longer work — because we are facing screen overload.

Screen overload refers not only to having too many screens but also to the different forms of screens, which means the scenarios behind them vary widely.

The same file may be used on a phone for checking messages, on a computer for more complex editing, and on a television for a meeting presentation. Complicating matters, these screens often belong to different brands of devices — you might be using a Honor phone, a Xiaomi TV, and an ASUS computer, with different forms, brands, and operating systems, and usually, you can only use a not-so-smart file transfer assistant… This is clearly not a good solution.

Futurists compare mobile smart devices to “organs” outside the human body, with information being the blood that connects these “organs” to the human body. The screen overload problem is essentially an issue of information overload and scenario overload. The development of the mobile internet has made atomic smart devices ubiquitous but has not completely solved the interconnectivity of information. Rather, the information that should flow freely has turned into isolated islands due to differences in device forms, brands, and systems.

To solve this problem, it is necessary to break free from the constraints of device forms, brands, and systems, starting from the bottom up.

Using AI to Connect Everything

Recently, Qualcomm unveiled its cross-platform technology Snapdragon Seamless at the Snapdragon Technology Summit. This solution aims to allow Snapdragon terminals using Android, Windows, and other operating systems to discover each other and work as if using a unified integrated system to share information. A few examples will clarify this —

When there’s a file to process on the phone, you can drag it over using the computer mouse:

Solving the Screen Island Problem with AI Connectivity

When listening to music with wireless headphones, if a meeting starts on the computer, the headphones will automatically pause the music and connect to the meeting:

Solving the Screen Island Problem with AI Connectivity

When doing yoga with AR glasses, the phone can track your body movements, and the information detected by the smartwatch will also be projected onto the lenses:

Solving the Screen Island Problem with AI Connectivity

The most interesting aspect is that in these scenarios, you do not need to specifically purchase devices from a certain brand — you can even use a Xiaomi phone, an ASUS computer, a Honor smartwatch, and a Vivo headphone to achieve all of this, because it is achieved through interconnectivity at the chip level, breaking the shackles of device forms, OEM brands, and operating systems in one go.

Support from underlying technology allows consumers to maintain maximum decision-making power. Dino Bekis, Vice President of Qualcomm Technology Company, stated:

Snapdragon Seamless breaks down the barriers between terminals and operating systems, embodying a truly “user-first” philosophy in cross-terminal solutions.

Facilitating the flow of overloaded information is not an easy task. Qualcomm’s solution is to introduce more advanced AI technologies and more powerful AI chips.

With the emergence of large language models like ChatGPT, AI has become a technology that everyone can feel, but to connect everything through AI, it must subtly integrate into daily life. Each person’s connected devices and scenarios are unique, and data models are highly personalized. Although cloud-based general large models have strong computing power, they have shortcomings in response speed and customization capabilities, and there are significant data privacy concerns.

This is also why device terminal manufacturers are exploring edge large models, and Qualcomm, as the largest underlying technology supplier, has the greatest first-mover advantage.

At the Snapdragon Summit, in addition to the performance-focused third-generation Snapdragon 8 mobile platform for Android flagship phones, Qualcomm also allocated significant space for the Snapdragon X Elite chip — a platform for AI PCs next year, employing a 4nm manufacturing process and surpassing similar products in energy efficiency, but more importantly, this platform is specially designed around AI capabilities.

0.6 seconds is the time it takes for a computer equipped with the Snapdragon X Elite to generate an image using Stable Diffusion. This open-source AI image generation model is the technical basis of the “Miao Duck Camera” which became popular at the beginning of the year for its ability to replace ID photos. If it were a cloud-based large model, you might have to wait several tens of seconds to generate an image, whereas the edge large model can do it in under a second. During the Snapdragon Summit, the Snapdragon X Elite chip was already able to run a large model with 13 billion parameters, making it one of the first platforms capable of running generative AI models with more than 10 billion parameters on the edge, the other being the third-generation Snapdragon 8.

Although there is still a gap compared to cloud-based large models like ChatGPT that run 100 billion parameters, with the development and iteration of the open-source ecosystem for generative AI models, the capabilities of generative models will become stronger, and the required parameters will decrease, ultimately becoming the most solid barrier for edge large models — because not everyone is willing to put their private data in the cloud.

Each year, half of the devices sold in the PC market are to business users, meaning many users do not want their business secrets transmitted online, and the advantage of edge large models lies in their excellent privacy, providing good customization capabilities while protecting privacy, making them more aligned with daily usage habits.

This is also why Qualcomm and Snapdragon chips emphasize the development of AI underlying capabilities and interconnectivity. When users switch to new terminals, Qualcomm can assist users in directly migrating personalized models to the new terminal, helping users achieve efficient work across different terminals through capabilities like Snapdragon Seamless, providing a seamless experience across terminals.

Moreover, the AI terminal assistants supported by Qualcomm will have the capability to access all underlying applications, allowing for the collection and organization of atomic information from phones and PCs across apps and terminals without needing to switch between multiple applications to complete complex tasks — the underlying technology research and development is driven by profound insights into the evolution of the mobile internet. This is something only manufacturers with chips deeply embedded in billions of mobile devices can achieve, and Qualcomm happens to be one of them.

At the Snapdragon Summit, we can see terminal manufacturers like Xiaomi, OPPO, and Honor, as well as operating system giants like Microsoft, and productivity software developers like Blackmagic Design. This Australian tech company is one of the most important “water sellers” in the film industry. They announced that their film editing software DaVinci Resolve will be available on the Snapdragon X Elite platform in 2024, representing the entry of advanced productivity “water sellers” — bringing together the creators of terminals, systems, and tools to discuss together; I think this is also the value of the Snapdragon Summit.

Undoubtedly, 2024 will be a year of mass emergence of AI devices and an important turning point for the replacement of smart devices. We will face more issues of coexisting with device connectivity, and a unified smart terminal ecosystem experience will be the most important product competitiveness, with Qualcomm and Snapdragon being the biggest support in this regard.

I think that by then, having more control over device selection will not be a bad thing.

Invested by the Father of ChatGPT, This Product Aims to Become the iPhone of the AI Era

Solving the Screen Island Problem with AI Connectivity

After a Week with the eSIM Version of iPad 10, It Changed My Work Style

Solving the Screen Island Problem with AI Connectivity

Leave a Comment Cancel reply