By Adam King, VP & General Manager of Personal Devices Business Unit at MediaTek. Follow Adam's newsletter to learn more about the latest Innovation Talk.
When we hear about generative AI, most of us think about using ChatGPT and similar platforms on our smartphones and laptops. What we don’t often think about is how our smart home devices are using generative AI to make our lives easier. Thanks to generative AI, the smart home is now a whole lot smarter, more personalized, and more intuitive.
Smart home assistants have been around for over a decade, so talking to a device feels normal by now. However, it’s easy to forget how far we’ve come even in the past few years. When these devices were first rolled out, you had to talk to them in a certain way. If you didn’t phrase a command just right, your device might not understand you. Generative AI has completely flipped the script. Now you can talk to your devices the same way you would talk to a person. That’s why I like to think of today’s smart home devices as personal assistants. Similar to a personal assistant, your devices will learn about your preferences and anticipate your needs.
Take the latest Amazon Echo devices as an example of how far this technology has come. These devices use generative AI to not only understand what you’re saying but also recognize who’s speaking. If you ask what’s on your schedule for today, you’ll get your personal rundown; when your spouse asks the same question, they will get a personalized answer as well. Additionally, with Alexa+ you can connect your device with tens of thousands of devices and services. Need to book a ride? Want a dinner reservation? Thanks to agentic AI, your device will take care of it all.
When you’re talking to your favorite smart home device, there’s a good chance it’s powered by MediaTek. We’re the No. 1 provider of chips for voice assistant devices and smart TVs, and we’ve specifically optimized our chips to process generative AI workloads at the edge. Some people wonder why edge processing is necessary when home Wi-Fi is so fast and reliable. In fact, a lot of the personal assistants on the market today use a hybrid of cloud and edge processing. However, edge processing still offers advantages for latency, efficiency, and privacy. We’re also seeing more companies design specialized small language models (SLMs) designed for specific tasks; these SLMs can run super efficiently and quickly on-device.
Edge computing is even more important for some other types of smart home devices like security cameras. If you were to send 4K video to the cloud 24/7, that would be a ton of data, especially if you have multiple cameras around your house. Processing that data on-device is so much more efficient—both from a power savings and cost savings perspective. A lot of consumers also appreciate the privacy benefits of processing information on-device. In the event of an incident, information can still be sent to the cloud as needed.
So what’s ahead for 2026? We’ll see personal assistant devices that can process even bigger AI models right on the device. More smart home gadgets will have built-in AI agents. And all your smart home devices will work together more seamlessly to make everyday life a little easier.
