Incorporating AI Into Portable Devices and What It Means for the End Consumer

9


During Barcelona’s Mobile World Congress (MWC) 2024 — one of the largest and most influential events for the “connectivity ecosystem” — George Zhao, the CEO of global technology brand and smart device maker Honor, delivered a keynote on the company’s latest products, technology and notably, how the company is harnessing the power of artificial intelligence (AI) for the next generation of its mobile and portable devices.

“Whether you like it or not, AI is coming into our daily lives,” he added in an interview with BoF after the keynote. Indeed, AI is poised to revolutionise portable devices. Honor is championing a “human-centric” approach — ensuring that all technological developments are created with the end user in mind, designed to “revolutionise the user experience with industry-leading AI capabilities, while breaking boundaries in human-device interaction,” Zhao said.

Honor’s products leverage both cloud-based AI and on-device AI. Honor’s expertise in on-device AI is enabled through its new operating system MagicOS 8.0, which fully integrates platform-level AI and the industry’s first intent-based user interface (IUI).

The Cloud AI allows for general-purpose applications of generative AI, completing functions such as simple question and answer services or text creation, even in an offline environment. The on-device AI predicts user behaviour with increased accuracy, from an intent-based shortcut recommendation feature, prioritising apps and functions for next-step usage based on learned behaviour, to an early phase of eye tracking — hands-free product usage enabled by tracking the user’s eye movement.

A crowd of people surrounding a central stage where a white Porsche car stands.

While the technology is still relatively nascent, by collecting and analysing data from users as well as from open-source models, the available use cases and technological capabilities are set to grow at speed. To optimise technological performance from its software and hardware, Honor funnels 11.5 percent of revenue back into research and development (R&D) and places a strong emphasis on “open collaboration” — working with leading businesses to drive the industry forward. Honor’s global partners for products and software include Intel, Microsoft, Google, Qualcomm and NVIDIA.

Crucially, the on-device AI is also designed to better protect and secure users’ personal data, ensuring personal information like home and work addresses or banking details remains within the device. This approach is aimed at alleviating consumer concern and distrust fuelled by uncertainty of where their data is ending up — Deloitte’s Connect Consumer Survey in 2023 found that 67 percent of smartphone users worry about data security and privacy on their phones, up 13 percentage points from 2022.

The company has also collaborated with Porsche Design for a special edition of its foldable mobile phone: the Porsche Design Honor Magic V2 RSR, which is set to enter the European market this month. It won the prestigious industry award ‘GLOMO Best in Show’ award for a device at the MWC 2024. Honor also teased that there will be a second jointly developed device coming soon based on the company’s AI-enabled bar phone Magic6 series, set to debut in March in China.

Porsche Lifestyle Group’s CEO Stefan Buescher said that the approach to AI with “human-centricity” is what “couples well with Porsche’s strict principles”. The companies worked together for over two years to “refine the functionality and the interface” of the phone, combining their respective expertise in design and technology.

Buescher added: “We are constantly aiming to infuse Porsche Design principles beyond automotive into every aspect of life, especially in an area dominated by technology. Our collaboration with Honor is set to redefine luxury with jointly crafted products that capture the essence of both brands and deliver an exceptional experience for our global customer audience.”

The following day at MWC 2024, Zhao featured on a panel discussion: ‘Putting Humans First in AI Development: Key Considerations for Creating AI for Smart Devices’. Moderated by John Hoffman, CEO of GSMA, Zhao sat alongside Alex Katouzian, the group general manager of the mobile, computer and XR wireless at technology software company, Qualcomm. Now, BoF shares key insights from the panel about the future of AI and how technology companies are catering to consumer needs.

Panel discussion at MWC 2024 on ‘Putting Humans First in AI Development: Key Considerations for Creating AI for Smart Devices’. Panellists on stage from left to right: Qualcomm's Alex Katouzian, Honor's George Zhao and GSMA's John Hoffman.

AI is shaping a cross-ecosystem experience and consumer product choices

George Zhao: In the smartphone industry, the media is [suggesting] that innovation has faced a bottleneck but from my point of view, it’s just time for the real innovation to boom. […] The opportunity for us is to reconstruct the operating systems with AI. Platform-level AI and [large language models, or] LLMs enable Honor to seamlessly blend AI devices such as AI PCs and AI smartphones into users’ daily lives, understanding their intent and anticipating their needs.

Today, we have a laptop, a PC, a smartphone, tablet, smartwatch — many devices, but they are [currently] isolated. [If you] want to transfer a photo from the phone to the PC, you need to first connect them together […] But in the future, this will be AI-enabled across devices. AI will reconstruct operating systems and the user experience and it can change [how we live] our lives.

Alex Katouzian: AI has been running in the background of phones for years now — the way you take a photograph in sunlight or low light, in darkness, the colour clarity of your pictures and your videos, your security algorithms, detecting malware, detecting false networks. You didn’t know you were using it, but it was there. The advancements we have made in AI in the past few years have been concentrated on how we can [improve] the user experience but without mentioning anything about AI.

Anything that [the consumer will] touch, if it has that AI capability, that intelligence, it’s going to transform their consumer experience.

AI will be a companion in every place that you are, meaning where you use personal devices — in the home, your car and at work. […] Intelligence in every device you interact with is where it is headed. […] It’s transformative. Anything that [the consumer will] touch, if it has that AI capability, that intelligence, it’s going to transform their consumer experience.

Now, AI is completely in the foreground. Consumers can pick out devices to say, “I’m going to rate this on its AI capabilities and what it can do for me, how much it can save time.” It’s become a central feature and function such as CPU speed or GPU speed. We want to make sure that technology and that assistance and time savings are drawn out as much as possible so the technology can be viewed by a consumer as something that they want [to] upgrade to.

AI should be human-centric and regulated to empower users

GZ: AI technology is developing so fast today and this makes people feel a bit scared, thinking AI will be more powerful than humans. From Honor’s point of view, we want the device to empower and enable people, to make them feel more confident.

[Being] “human-centric” is our strategy and philosophy, so when we design, people are [considered] in [every] step. Through our efforts, on-device AI shall co-exist in harmony and will work together with Cloud AI, so everyone can enjoy an AI life and humans will control everything.

Consumers need time [to build trust]. They need to build trust from the small things to everything and […] we must show from our side that we follow [our] principles [of privacy first]. So, in anything and everything we do, we must gain their trust. Trust in the AI era is life and we must protect that.

AK: About one year ago at [MWC], we showed the first AI model, Stable Diffusion, running on a phone without any connection to any [wireless] network, taking words as input and it would draw a picture. That took 20 seconds to run, it had no connection to the Cloud and it was running an LLM, an LVM (a visual model) that went from words to pictures. About nine months later, we showed the same Stable Diffusion running in half a second.

[These capabilities] could be looked at as scary, but I look at it as: innovation will bring goodness and better user experience to people. All of us in this field adhere to your privacy and your security and we have to make sure things are regulated so it doesn’t get out of hand. However, I think that innovation will really excite people.

[To build trust,] we work with our partners to translate this technology into a language everyone can relate to and understand. In terms of security, we talk about encryptions, private space, we talk about only accessible by the user with their own password and the message [can be understood] properly. But as soon as we start using big words in technology, all of us get lost.

Privacy and security should be optimised through hybridised AI

GZ: Honor follows the principle PFAST — Privacy, Fairness and Justice, Accountability, Security and Reliability, Transparency and Controllability. The philosophy is that all data belongs to the end consumer. You cannot manage consumer data without their permission and that is a strict principle of on-device AI.

If you have AI on your phone, you can ask the AI to navigate to your home to pick up your son, then go to your wife’s workplace to pick her up, then go to your favourite restaurant for dinner. But this means the AI must know your home address, your wife’s office address and the address of your favourite restaurant. That includes a lot of private information. With on-device AI, we can protect that.

The philosophy is that all data belongs to the end consumer. You cannot manage consumer data without their permission

When we deliver an advanced user experience, we must [ensure] the user data is secure and well protected. The smartphone device must protect the user data and that is why we implement on-device AI on our devices as well as collaborate with Cloud AI.

Cross-collaboration and interconnectedness is critical to industry-wide success

AK: We want to follow open standards; we want to provide capabilities to developers; we want to make it easy for people to develop and innovate on top of our platforms. So, cooperation between our partners is key and we believe multiple smart brains are better than one. In some cases, fragmentation is bad for our industry, but cooperation and openness are actually very good. […] As long as we can continue to innovate and cooperate with our partners, we are boundless.

GZ: Today, we have a lot of choices for electronics. But soon, across the different ecosystems — Windows, Android, the LT ecosystem — data could collaborate across these with AI capabilities. With [the likes of] Qualcomm and Google as our partners, we can make services migrate across the different operating systems and devices.

From my point of view, when we [the industry] work together, we can bring a fantastic and seamless user experience for everybody. Don’t build a walled garden — be open to collaboration. […] When we work together, we can share the AI capability. Then we can work with Google, Microsoft — we can break across operating systems and build a seamless user experience. The boundary is not in the technology and hardware part but on the mind’s side.



Source link

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More