How Gemini will ‘improve’ Google Assistant for speakers, displays

Abner Li | Nov 15 2024 - 2:46 pm PT

Back in August with the Google TV Streamer and new Nest Learning Thermostat, Google revealed that Assistant is being upgraded with Gemini.

About APK Insight: In this “APK Insight” post, we’ve decompiled the latest version of an application that Google uploaded to the Play Store. When we decompile these files (called APKs, in the case of Android apps), we’re able to see various lines of code within that hint at possible future features. Keep in mind that Google may or may not ever ship these features, and our interpretation of what they are may be imperfect. We’ll try to enable those that are closer to being finished, however, to show you how they’ll look in case that they do ship. With that in mind, read on.

With the Gemini large language models, Google wants to make Assistant more “natural and helpful.” Questions and follow-ups don’t have to be structured in a precise manner. Generative capabilities allow for more back-and-forth interactions about a wider variety of topics.

Google is also using LLMs to improve core experiences like media playback and setting timers. We already have a test of this with the Google Home Gemini Extension on Android allowing for more natural queries that don’t have to explicitly reference devices.

Google Home 3.26.1.3 contains an introduction prompt that users will see once this “improved Assistant” is live for speakers and smart displays:

<string name=”gia_title_text”>Introducing an improved Assistant</string>

Richer responses: Hear complex topics summarized quickly and clearly
New voices: New, more natural voices make it easier to interact with your speakers and displays [Previous clip below]
Better understanding: Pause, ask multiple questions, change your mind, and you’ll still get the help you need

This welcome message is not yet live, though Google this week started testing Gemini-powered AI Camera Search and descriptions. A broader launcher is expected in 2025.