Quick Links
In recent years, neural processing units (NPUs) have become more and more capable. With these new AI edge processors becoming more accessible to manufacturers, we are now experiencing a surge of new innovative AI products that can revolutionize how we use and interact with technology. Here are the ones I’m really excited about:
1 AR Headsets and Smart Glasses
AI is bringing wearables into the future, and AR glasses and headsets are leading the way. Meta’s Quest 3, for example, combines its own Meta AI assistant with miniaturized hardware, including advanced sensors, NPUs, and AI-enhanced pass-through cameras. This allows users to use the Meta Quest 3 to enjoy mixed-reality gaming and virtual workout sessions, watch virtual content online, and do daily productivity tasks while being aware of their physical surroundings.
Apple users may also opt for Apple’s Vision Pro AR headset, which supports and maximizes the use of other Apple devices. The Vision Pro integrates AI with highly sophisticated sensors and LiDAR technology that map your surroundings to merge the physical and digital worlds into one seamless environment. This allows users to enjoy an immersive VR experience similar to that of the Meta Quest 3 with the addition of the Apple ecosystem.
Looking for smart glasses to take on your travels? Ray-Ban smart glasses now integrate Meta’s AI-powered assistant. Pair that with a 12MP camera, auto-transition lenses, earphones, and an AI edge processor, and you get a cool AI-augmented device you can bring anywhere for hand-free photo and video recording, auto-language translations, voice commands, and live notifications.
2 AI Smartphones
Thanks to chips like Google’s Tensor and Apple’s Bionic series, smartphones have evolved into AI powerhouses. These System on a Chip (SoC) processors house neural processing units such as Google’s Tensor Processing Unit (TPU) and Apple’s Neural Engine (ANE). These chips handle everything from FaceID and augmented reality apps to computational photography by distributing AI workloads efficiently across the CPU, GPU, and dedicated neural processors.
Google’s Tensor chip excels in computational photography and natural language processing. It provides the newest Pixel phones with some of the most useful features, such as AI-assisted night mode, local AI image editing, real-time transcription and translation, song detection, and adaptive user experience.
On the other hand, Apple’s latest Bionic chips provide on-chip AI capabilities in every iPhone, such as enhanced augmented reality experiences, hands-free voice commands with Siri, quick and secure facial recognition, smooth gaming with AR apps, and voice recognition features.
3 Home Robots
AI-powered home robotics is an area where advancements in sensors, machine learning, and autonomous navigation come together. New AI-driven home robots such as Amazon’s Astro utilize advanced AI for facial recognition and machine learning to navigate your home autonomously.
What makes Astro stand out is its fusion of AI and sensor technologies like LIDAR and SLAM (Simultaneous Localization and Mapping), which allow it to create a map of your home and detect obstacles. Its AI enables it to move smoothly and understand contextual cues, such as recognizing faces, interpreting voice commands, and learning your household preferences over time.
With home robots becoming increasingly popular, we could potentially use them for home security (monitoring), moving smart assistants, a control hub for smart home devices, and robot pets offering companionship for children, the elderly, or the disabled.
4 AI Laptops
With AI becoming an integral part of work, entertainment, and regular day-to-day tasks, AI-powered laptops are now being released to the public. Similar to AI smartphones, these laptops feature NPUs as part of their SoCs. Some of the more notable AI processors include Intel’s Lunar Lake, AMD’s Strix Point, Qualcomm’s X Elite, and Apple’s M series of processors, each designed to harness AI’s power in different ways.
Intel’s Lunar Lake processors promise to leverage AI to optimize energy efficiency, predict user behavior, and intelligently manage system resources. With AI-reliant software such as the upcoming Microsoft Copilot expected to become a standard for Windows systems, a Lunar Lake AI processor should help further decrease battery usage, with its integrated NPU efficiently handling most of the tasks.
While Intel focuses on power efficiency, AMD’s Strix Point processors push AI workloads to the forefront of everyday computing tasks, such as enhancing gaming, video conferencing, and multimedia content creation. Its use of AI accelerators allows for effective real-time noise reduction, gesture recognition, and video editing optimizations, making Strix Point the go-to processor for high-performance laptops suitable for gaming and content creation.
A rather unexpected but welcome addition to the latest AI processors is Qualcomm’s Snapdragon X-Elite processors. Its ARM architecture makes the X-Elite inherently power efficient but at the cost of limited software support. However, as Windows in ARM becomes more popular, ARM laptops seem to have a bright future. Not only are they power efficient, running ARM native apps should also make them capable devices for gaming and other resource-intensive tasks.
Another ARM-based AI processor is Apple’s M-series of chips. They include AI Neural Engines that dramatically increase the speed and efficiency of image and video processing tasks. The unified memory architecture in these chips also ensures that machine learning models have direct access to high-speed memory, enabling faster AI-driven applications like voice dictation and augmented reality.
5 AI Cameras
AI-powered cameras are transforming how we capture moments and produce content. This new breed of camera makes it easier to capture smooth, professional-level footage without complicated manual settings and setups.
For instance, the Hover Air X1 drone is equipped with AI that can autonomously follow subjects, avoid obstacles, and optimize framing for the perfect shot. Another newly released AI drone is DJI’s Neo, which takes framing a step further by employing reinforcement learning, allowing the drone to predict movement patterns and adjust its flight paths dynamically. Both drones’ high levels of autonomy are only made possible by deep learning algorithms and real-time recognition software.
AI cameras can also be found in actions in handheld cameras like Insta360’s X4 action camera. Its AI-enhanced stabilization, subject tracking capabilities, and voice commands use the 360 platform efficiently to capture the surroundings while keeping you in focus the entire time.
Stationary cameras such as the Insta360 Link webcam also use AI to conveniently track and follow a subject with its AI camera and motorized tilt and pan functions. Its whiteboard mode uses AI to focus on a whiteboard and frame it to look perfectly cropped and leveled for the audience to see. Other advanced features, such as intelligent noise cancelation, gesture control, and overhead mode, make the Inta360 Link the perfect companion for online classes and presentations.
All of this is made possible by onboard AI accelerators, which allow these devices to perform complex computations without relying on cloud-based processing, ensuring quick response times and real-time decision-making.
6 Self-Driving Cars
Lastly, we have self-driving cars. Self-driving technology has been around for quite some time now, but it’s only been in recent years that AI has helped self-driving vehicles to get past conditional automation (level 3) to high automation (level 4) based on the 5 levels of driving automation.
Waymo’s fully autonomous technology takes the lead in self-driving cars. They use AI for tasks like obstacle detection, route optimization, and complex urban navigation. Waymo’s cars rely on deep learning and powerful processors to achieve full autonomy. Crucially, the AI is trained on millions of miles of real-world driving data, making it safer and smarter over time.
Tesla’s Autopilot takes AI deep learning even further by only using machine vision to enable highly autonomous driving on its latest Cybertruck. What powers Tesla’s self-driving capabilities is the in-house developed Full Self-Driving (FSD) chips, which contain a dedicated AI neural net to process sensor data in real-time without a cloud connection.
The integration of AI into everyday products is transforming how we interact with technology. From wearables like AR glasses to smartphones, laptops, and even home robots, AI enhances functionality and user experience in ways we could only dream of a few years ago. As these technologies continue to improve, we can expect an even brighter future for AI-enhanced technology.
