How Google plans to fix its smart home mess with Gemini

The original promise of smart speakers was to revolutionize our daily lives by shifting tasks from manual input on our phones and computers to simple, hands-free voice commands. However, they largely failed to live up to this promise. In fact, some smart speakers, including those in Google’s Home ecosystem, have even seen a significant decline in reliability.

This decline raises the question of what, if anything, tech companies are doing to reverse the trend. Google’s answer is to bring its Gemini AI into the smart home. Unveiled this week, Gemini for Home upgrades the conversational experience in Google Home smart speakers, displays, doorbells, and cameras with AI, while also introducing smarter media controls, improved alerts, and more intelligent search capabilities.

Google’s announcement certainly sounds like the start of something big, with Gemini for Home’s capabilities aligning more closely with the platform’s original pitch. However, after months of dealing with glitches and bugs, it’s tough to ask users to trust that this time will be different. Many say they’re done with Google’s smart home ecosystem, which is understandable after investing hundreds, if not thousands, of dollars and countless hours into a system that barely works.

Google clearly has a lot of work cut out for it if it wants to win back the trust of these users, and the company is aware of this. I recently interviewed Anish Kattukaran, Chief Product Officer at Google Home and Nest, about the Gemini for Home launch and how the company plans to win back trust in its ecosystem.

Kattukaran began by acknowledging the skepticism from long-term users, admitting to past performance issues with Google Home and the Google Assistant. He mentioned how the team has taken the public’s feedback to heart and has been more actively engaging with the community on forums like X and Reddit.

He explained the company faced a tough choice: either continue applying “band-aid” fixes to the decade-old Google Assistant technology or build for the future with Gemini. While many existing users would be content with the former, doing so would only prolong the stagnation in the smart home space. The introduction of large language models like Gemini finally provides the necessary intelligence to significantly improve the smart home experience, which is ultimately why Google decided to rip the band-aid off and make the transition.

However, Google knows that people have invested lots of time and money into its ecosystem and that it would be asking a lot for them to buy new hardware. That’s why Google is bringing Gemini for Home to every speaker, smart display, camera, and doorbell made in the last decade. This required rewriting millions of lines of code to migrate this older hardware to the new foundation, according to Kattukaran. Unfortunately, some older Nest thermostats built on obsolete backends didn’t make the cut.

Gemini for Google Home won’t just bring new functionality; it’ll also improve the foundational experience, says Kattukaran. It will handle the fundamental tasks people expect — such as media controls, smart home commands, timers, and alarms — but will enable them with more natural, less formulaic commands. Furthermore, the team has worked to significantly improve the performance and stability of the Google Home app, reducing crashes and speeding up load times for cameras and other devices.

These improvements will go a long way towards rebuilding trust in Google’s smart home ecosystem, but trust isn’t earned in a single day. Kattukaran stressed that his team wants to re-earn trust over time through concerted action. He pointed to the team’s multi-year effort to migrate all its Nest devices to the Google Home app as a sign of its commitment and promised to release updates and improvements more frequently. The best proof of Google’s commitment will be seeing the company actively roll out these improvements while maintaining transparency with its users.