As the adoption of generative artificial intelligence accelerates, more computing will be done in the hands of end users—literally. Increasingly, AI will be embedded in consumer devices such as smartphones, notebooks, wearables, automobiles and drones, creating new opportunities and challenges for the manufacturers of these devices.
Generative AI’s phenomenal capabilities are power-intensive. So far, the processing needed to run sophisticated, mainstream generative AI models can only take place in the cloud. While the cloud will remain the foundation of AI infrastructure, more AI applications, functions and services require faster or more secure computing closer to the consumer. “That’s driving the need for AI algorithms that run locally on the devices rather than on a centralized cloud—or what’s known as the AI at the Edge,” says Ed Stanley, Morgan Stanley’s Head of Thematic Research in London.
By 2025, Edge AI will be responsible for half of all enterprise data created, according to an estimate by technology market researcher Gartner Inc. While there are many hurdles to reaching commercial viability, the opportunity to tap into 30 billion devices could reduce cost, increase personalization, and improve security and privacy. In addition, faster algorithms on the Edge can reduce latency (i.e., the lag in an app’s response time as it communicates with the cloud).
“If 2023 was the year of generative AI, 2024 could be the year the technology moves to the Edge,” says Stanley. “We think this trend will pick up steam in 2024, and along with it, opportunities for hardware makers and component suppliers that can help put AI directly into consumers' hands.”
New Smartphones Lead the Charge
Smartphones currently on the market rely on traditional processors and cloud-based computing, and the only AI-enabled programs are features like face recognition, voice assist and low-light photography. Device sales have slowed in recent years, and many investors expect that smartphones will follow the trajectory of personal computers, with multi-year downturns as consumers hold onto their devices for longer due to lack of new features, sensitivity to pricing and other factors.
But thanks in part to Edge AI, Morgan Stanley analysts think the smartphone market is poised for an upswing and predict that shipments, which have slowed since 2021, will rise by 3.9% this year and 4.4% next year.
“Given the size of the smartphone market and consumers’ familiarity with them, it makes sense that they will lead the way in bringing AI to the Edge,” says Morgan Stanley’s U.S. Hardware analyst Erik Woodring. “This year should bring a rollout of generative AI-enabled operating systems, as well as next-generation devices and voice assistants that could spur a cycle of smartphone upgrades.”
However, the move to the Edge will require new smartphone capabilities, especially to improve battery life, power consumption, processing speed and memory. Manufacturers with the strongest brands and balance sheets are best positioned to take the lead in the hardware arms race.
Killer Apps
In addition to hardware, AI itself continues to evolve. New generations of AI models are designed be more flexible and adaptable for a wide range of uses, including Edge devices. Other beneficiaries include smartphone memory players, integrated circuit makers and camera parts suppliers that support new AI applications.
What can you expect from your phone in the next year?
- “Always-sensing cameras” that automatically activate or lock the screen by detecting if the user is looking at it without the need to touch the screen. This feature could also automatically launch applications such as online payment and food ordering by detecting bar codes.
- Gesture controls for when the user is unable to hold their devices, such as while cooking or exercising.
- Desktop-quality gaming experiences that offer ultra-realistic graphics with cinematic detail, all with smoother interactions and blazing-fast response times.
- Professional-level photography in which image processors enhance photos and video in real time by recognizing each element in a frame—faces, hair, glasses, objects—and fine tune each, eliminating the need for retouching later.
- Smarter voice assistance that is more responsive and tuned the user’s voice and speech patterns, and can launch or suggest apps based on auditory clues.
“With Edge AI becoming part of everyday life, we see significant opportunities ahead as new hardware provides a platform for developers to create ground-breaking generative AI apps, which could trigger a new hardware product cycle that liftsservices sales,” says Woodring.
For deeper insights and analysis, ask your Morgan Stanley Representative or Financial Advisor for the full reports, “Tech Diffusion: Edge AI—Growing Impetus” (Nov. 7, 2023), “Edging Into a Smartphone Upcycle” (Nov. 9, 2023) and “Edge AI: Product Releases on Track, But Where Are Killer Apps?”
