YouTube Lets Users Build Their Own Algorithm with AI-Powered Custom Feeds

YouTube Introduces AI-Powered Custom Feeds for Personalized Content Discovery

YouTube has officially announced Custom Feeds, an AI-powered feature that allows users to build their own algorithm by describing the kinds of videos they want to see in natural language1

1

2

2

. The feature leverages Google's Gemini AI to create personalized video recommendation feeds based on user prompts, giving viewers more control over recommendations and addressing the platform's tendency to overreact to single video views3

3

.

Users can type descriptions into a prompt box to customize their algorithm, whether requesting "evening wind down with long documentaries," "new DIY creators," or "relaxing commentary videos to unwind with"1

1

4

4

. The AI-driven personalization allows for detailed, multi-layered prompts where users can specify what to include, exclude, and prioritize in their content curation1

1

.

How Custom Feeds Work and Where They Appear

Once created, custom feeds are pinned to the top of the YouTube home page in their own tabs, accessible via a "Your custom feeds" button under the search bar2

2

. Users can create and save multiple Custom Feeds, each serving different purposes—one for morning workouts, another for commute podcasts, or a third for late-night relaxation3

3

. The feeds automatically refresh with new recommendations and can be edited anytime2

2

.

The AI-driven feed customization won't replace YouTube's main feed but will complement it, allowing users to toggle between their personalized recommendation feeds and the standard algorithm-driven home page4

4

. This separation keeps different interests in their own lanes, preventing one random video from hijacking the entire recommendation system3

3

.

YouTube Joins Growing Trend in Algorithm Personalization

YouTube's move follows similar initiatives across social networks where users can build their own algorithm. Bluesky pioneered this approach in 2023 as a differentiation strategy from rivals like X, later rolling out Attie, an AI tool that simplified feed-building using natural language1

1

4

4

. Meta followed suit with custom feed options on Threads and Instagram's "Your Algorithm" tab, while Spotify recently launched Taste Profiles for Premium subscribers2

2

.

Mark Zuckerberg explained Meta's vision during an earnings call earlier this year: "Today, our apps feel like algorithms that recommend content. Soon, you'll open our apps, and you'll have an AI that understands you and also happens to be able to show you great content or even generate great personalized content for you"2

2

. This shift toward AI-driven personalization reflects platforms betting that hyperpersonalized feeds will drive greater user experience and engagement2

2

.

Navigating YouTube's 20 Billion Video Corpus

Emily Moxley, VP of Product Management for Viewer AI at YouTube, emphasized the feature's purpose in helping users explore the platform's vast content library. "There's over 20 billion videos in the YouTube corpus, so you're always one search away from a treasure trove of topics," Moxley explained1

1

4

4

. Custom Feeds aim to make this exploration more intentional and less dependent on algorithmic guesswork3

3

.

The feature expands on a limited test YouTube conducted last year, where select users could tell an AI chatbot what they wanted to see via a "Your Custom Feed" chip beside the Home tab3

3

. The new version broadens this capability, allowing multiple saved feeds rather than a single experimental option3

3

.

Rollout Timeline and Platform Availability

YouTube announced that support for creating multiple custom feeds will roll out starting October, initially to users in the United States1

1

4

4

. The feature will be available across YouTube on web, mobile apps, and TVs3

3

. The announcement came during YouTube's Made on YouTube event, which also introduced creator tools including conversational editing, video A/B testing, and improved likeness detection3

3

.

While the effectiveness of Custom Feeds may depend on users' skills at prompt engineering, the feature represents a significant shift toward giving viewers agency over their content consumption4

4

. As platforms increasingly rely on AI to understand and serve user preferences, YouTube's Custom Feeds could reshape how viewers interact with the world's largest video platform and set expectations for more control over recommendations across all social networks.