Basic phone assistant functionality Gemini should have had from day one
In the ongoing effort to make artificial intelligence feel genuinely useful in daily life, Google has extended its Gemini assistant to control YouTube Music through natural voice commands — a capability its predecessor handled years ago. The rollout, following the I/O 2024 announcement, marks a quiet admission that Gemini launched incomplete, and that the work of becoming a true personal assistant is being assembled piece by piece. It is a small but telling moment in the larger story of how AI platforms must earn the trust of ordinary habit.
- Gemini could hold a conversation but couldn't play a song — a gap that made it feel less capable than the assistant it was meant to replace.
- The YouTube Music extension is now live, letting users summon tracks, artists, playlists, and even lyrics-based searches through natural language on Android and iOS.
- A friction point lingers on the web version, where an extra click is required to actually open the music player, while mobile integration runs more smoothly.
- Google has acknowledged the broader incompleteness, with Calendar, Keep, Tasks, and Clock extensions promised in the months ahead.
- The slow, app-by-app rollout raises a harder question: whether Google has a unified framework for third-party integration, or is simply patching holes one at a time.
Google is rolling out a YouTube Music extension for Gemini, its AI assistant — bringing what should have been a foundational feature into existence well after launch. Announced at Google I/O in May 2024, the extension is now broadly available. Users connect their YouTube Music account at gemini.google.com/extensions, after which Gemini can respond to a wide range of music requests.
On Android, commands like "Play [song]" or "Play [artist]" trigger background playback immediately. On iOS, the YouTube Music app opens to complete the request. In both cases, Gemini surfaces a card with cover art, artist name, duration, and play count. The extension handles everything from simple playlist triggers — "Play music that I like" pulls up your Liked Music playlist — to more expressive requests like finding songs by lyrics, starting a radio station, or surfacing artists keeping classic blues alive. English-only for now, and podcasts on YouTube are supported as well.
The web experience introduces a small but notable friction: the card appears, but opening the actual player requires an extra click into a new tab. Mobile feels more natural.
What makes this rollout quietly significant is what it reveals about Gemini's origins. Google Assistant could do this years ago. That Gemini is only now catching up underscores how much was missing at launch — and how far the platform still has to go before it can genuinely replace its predecessor. Google has signaled that Calendar, Keep, Tasks, and Clock integrations are coming in the months ahead, each one another step toward a complete assistant. But the one-at-a-time pace suggests the company is still building the bridge, not yet opening the road.
Google is rolling out a YouTube Music extension for Gemini, its AI assistant, bringing a capability that should have existed months ago: the ability to play a song by asking for it. The extension, announced at Google's I/O developer conference in May 2024, is now available more widely. Users can activate it by visiting gemini.google.com/extensions and connecting their YouTube Music account.
Once connected, Gemini gains access to your music library and can respond to straightforward commands. On Android, saying "Play [song]" or "Play [artist]" will start playback in the background. iOS users see the YouTube Music app open to complete the request. Either way, Gemini displays a card showing the cover art, artist name, song duration, and play count—the basic metadata you'd expect.
The extension understands a range of prompt formats, from the simple ("Play music that I like" triggers your Liked Music playlist) to the more specific. You can ask it to find a song by artist name, locate an album, search by genre, identify a track from lyrics, pull up a music video, or start a radio station based on a song. More conversational requests work too: "Show me some rock music," "Turn on White Noise," or even "Show me songs from modern artists keeping classic blues alive." For now, Google limits the feature to English-language prompts.
There's a friction point depending on where you're using it. On the Gemini website, the extension shows you the card but requires an extra click to open the actual music player in a new tab. On mobile, the integration is smoother. The extension also works with podcasts already available on YouTube, expanding its usefulness slightly beyond music.
This is, fundamentally, basic phone assistant functionality—the kind of thing Google Assistant could do years ago. That Gemini is only now getting it underscores how incomplete the AI assistant felt at launch. If Google wanted Gemini to genuinely replace Google Assistant, this kind of media control should have been built in from the start. The gap points to a larger problem: there's no straightforward way to integrate Gemini with most of the apps and services on your phone.
Google is aware of the gap and moving to fill it. In the coming months, Gemini will gain extensions for Calendar, Keep, Tasks, and utilities like Android's Clock app. Each addition is a step toward making Gemini feel like a complete assistant rather than a conversational search engine. But the slow rollout also suggests the company is building these integrations one at a time, rather than having a unified framework that would let developers plug in their own apps quickly. For now, YouTube Music is the latest piece of the puzzle.
Notable Quotes
This is basic phone assistant functionality that Gemini should have had from day one if it wanted to replace Google Assistant.— 9to5Google reporting
The Hearth Conversation Another angle on the story
Why did it take Google this long to let Gemini control YouTube Music? Didn't Google Assistant do this years ago?
Yes. Google Assistant could play music on command for a long time. Gemini launched without that basic capability, which was a real gap. It's a reminder that Gemini wasn't built as a direct replacement for Assistant—it was built as a conversational AI first, and the assistant features came later.
So this extension is Google catching up to where they already were?
Partly, yes. But it's also a different architecture. Instead of baking everything into the assistant, Google is building Gemini as a platform with pluggable extensions. That's more flexible long-term, but it means starting from scratch on integrations.
Does the extension actually work well, or is it clunky?
On mobile it's smooth—you say "Play this song" and it plays. On the web version, you get a card but have to click through to actually start the music. It's functional but not seamless.
What about the conversational commands? Can you really ask it to "show me songs from modern artists keeping classic blues alive"?
Yes, and that's where it gets interesting. It's not just command-and-control. Gemini can parse more natural language. But it only works in English right now, which is a real limitation.
What comes next? Is Google going to do this for every app?
They're planning extensions for Calendar, Keep, Tasks, and the Clock app. But the pace suggests they're building these one by one. There's no indication yet that third-party developers can easily build their own extensions.