HaoTech
After years of declining sales, mobile phone manufacturers are placing their bets on artificial intelligence. Recent announcements indicate that embedded AI will significantly change the way we use our phones, interact with applications, and even how we drive.
With features such as conversational commerce, advanced text generation, and in-image search becoming more prevalent, the era of the basic ‘smart’ phone may be drawing to a close.
What does this mean for consumers? Can AI-enabled phones truly live up to their promise?
A wave of embedded AI announcements from companies like Samsung suggests that 2024 will be the year of the AI smartphone.
In this piece, Techopedia delves into the latest trends in embedded AI and speaks with experts about what the future holds.
Key Takeaways:
- The AI revolution has disrupted software, search, and SaaS sectors. Now, it’s targeting mobile.
- Manufacturers are hoping that embedded AI will reignite interest in the category and shorten upgrade cycles.
- While some of the current features might seem gimmicky, analysts believe a new era of handheld computing is on the horizon.
- Addressing privacy and security concerns is a crucial consideration for everyone involved in the smartphone AI ecosystem.
AI Smartphones Making Headlines
At the Mobile World Congress (MWC) in Barcelona this February, AI was the hot topic, with major tech giants like Samsung, Qualcomm, and Google heralding a new era of AI at your fingertips. Samsung was at the forefront, unveiling its Galaxy S24 line of AI-powered phones in January.
Brands like Xiaomi and Honor demonstrated the generative AI features of their latest devices, whereas Qualcomm and Mediatek introduced chips designed for on-device AI processing.
While high-end smartphones have incorporated AI functionalities for some time, previously, much of this processing occurred in the cloud.
The latest wave of AI smartphones boasts the necessary hardware, storage, and operating system capabilities to perform intensive computational tasks directly on the device. This advancement enables quicker, more responsive, and highly personalized user experiences.
The Biggest Embedded AI Announcements So Far in 2024
Samsung AI Smartphone S24
In January, Samsung’s launch of the S24 series positioned the Korean tech giant as the pioneer of the ‘first AI smartphone’.
Leveraging Google’s Gemini AI and the Imagen 2-based text-to-image tool, the trio of S24 series smartphones are equipped with an on-device Large Language Model (LLM) that seamlessly integrates with the Android 14 OS.
This integration extends AI functionalities into Samsung’s own suite of applications, notably incorporating advanced photo-editing features through Imagen 2.
Xiaomi 14
Introduced in October and officially launched thereafter, the Xiaomi 14, alongside its high-end counterpart, the Xiaomi 14 Ultra, garnered significant attention at the Mobile World Congress (MWC).
Both models are the first smartphones to offer on-device generative AI capabilities, powered by Qualcomm’s Snapdragon 8 Gen 3 platform.
Key features include the innovative ability to expand an image beyond its original boundaries, with AI intuitively filling in the extended space. Additionally, users can train the AI with their facial images to generate ‘selfies’ that imaginatively place them in new or exotic locales.
Honor Magic 6
The Honor Magic 6 Pro, unveiled in Barcelona, introduces users to an AI-powered navigation feature, enabling control of the phone through eye movements. It includes a tracking capability that observes where you’re looking and anticipates your next desired action.
The innovative feature, as claimed by the Chinese manufacturer, even extends to the possibility of controlling your car with just your eye movements, showcasing a new level of interaction between AI technology and everyday devices.
Qualcomm
Executing an AI model on a device demands substantial processing capability. Qualcomm introduced the first AI-ready mobile hardware towards the end of last year and at the Mobile World Congress (MWC), it unveiled new chips equipped with inherent AI functionalities.
The Snapdragon X80 5G modem, specifically designed for AI smartphones, tablets, PCs, and other devices, supports native AI operations. Its specialized AI processor is engineered to enhance data speeds, reduce latency, and optimize the transition across different wireless spectra, paving the way for a new era of connected devices with advanced AI capabilities.
Google attended the Mobile World Congress (MWC) with the ambition of establishing Android as the premier platform for mobile AI.
The company showcased new features for the operating system in Barcelona, such as the integration with Gemini, which allows users to directly engage with Gemini AI from within the Google Messages app on any Android device.
Additionally, an upcoming feature will enable the summarization and auditory playback of lengthy texts and bustling group chats, particularly useful while driving, enhancing safety and convenience for users on the move.
AI Smartphones Momentum: Exploring the Current Shift
The International Data Corporation (IDC) Worldwide Quarterly Mobile Phone Tracker indicates a 3.2% decline in global smartphone sales, totaling 1.17 billion units in 2023. High-end smartphones have achieved a plateau in quality and capabilities, leading to a market where differentiation is increasingly challenging. Consumers often respond with indifference to new releases, underlining the industry’s need for a revitalizing element. Embedded AI is seen as a potential catalyst for innovation and growth within the sector.
However, the question remains: Is this expectation realistic? Runar Bjørhovde, an analyst at Canalys, shared with Techopedia that the ultimate AI smartphone “super use case” has yet to emerge. He points out that smartphones have utilized a form of small language model for years, aiding in text suggestions and photo enhancements.
For a smartphone to transition to a fully “AI capable” device, it would need to support a comprehensive Large Language Model (LLM) like ChatGPT natively, without compromising on speed or performance. The latest generation of devices seems to pave the way for this advancement.
The launch of ChatGPT took many hardware manufacturers by surprise, highlighting an upcoming paradigm shift. This shift may offer new opportunities for customer monetization and market differentiation through AI.
Canalys defines an AI smartphone as one that can run an LLM like ChatGPT natively, supported by hardware, storage, and OS improvements that allow for swift processing and detailed results. Despite the current AI features—mainly focused on image modifications—being considered gimmicky, the research firm predicts that AI-capable smartphones will constitute 45% of the market by 2027.
Anshu Goel, the Lead Analyst of Digital Consumer at GSMA Intelligence, supports the perspective that the appreciation for advanced AI among smartphone users is expected to grow as this technology gradually integrates across a broad range of mobile digital services.
Goel further mentions that GSMA’s research reveals nearly half of the users intending to upgrade their phones regard AI features in smartphones as “very or somewhat important.” This insight underlines the growing significance of AI capabilities in the decision-making process for consumers looking to purchase new handsets, suggesting a shift in consumer expectations towards more intelligent and personalized mobile experiences.
Security Concerns: The Primary Challenge for Embedded AI
The integration of AI into smartphones could potentially open new avenues for cybercriminals. Techopedia’s Linda Rosencrance highlighted insights from security consultant Frank Abagnale, who was depicted by Leonardo DiCaprio in the film Catch Me If You Can. Abagnale believes that AI capabilities on smartphones could lead to more complex and cunning social engineering scams.
To counteract these emerging threats, Abagnale recommends a shift away from traditional password security towards the adoption of anonymized passkeys for enhanced protection of online activities. He shared his extensive experience, noting, “For the past 20 years, I have been a member of the Board of Advisors for a technology company that developed what is known today as passkey technology. This technology has been embraced by the FIDO Alliance, as well as tech giants like Apple, Google, and Microsoft. Now, passkey technology is a feature on every device, marking a significant move towards eliminating usernames and passwords. This change is both significant and long overdue.” This perspective underscores the urgent need for evolving security measures in the age of AI-enhanced mobile technology.
The Bottom Line
The introduction of AI into smartphones might open new doors for cybercriminals. Security expert Frank Abagnale highlights the potential for more complex scams with AI, suggesting a move to passkey technology for better security, a method now supported by major tech companies.
The Outlook on AI Smartphones
While the promise of AI in smartphones suggests a shift away from traditional models, skepticism remains due to the preliminary nature of many features. Nonetheless, innovative technologies like eye-tracking controls and app-less AI concepts hint at significant potential, awaiting realization.
