What YouTube’s AI Custom Feed Is and Why It Matters
YouTube’s AI custom feed is a new “Your custom feed” feature that lets you describe what you want to watch in natural language, then automatically builds a personalized video playlist-style feed you can pin, refresh, and refine over time without relying only on the usual recommendation algorithm. Instead of passively accepting whatever appears on your Home page, you now type a prompt about your interests, mood, or goals and get a tailored stream of videos that fits that description. This AI feed builder sits alongside your standard Home feed rather than replacing it, which means you can switch back and forth whenever you like. For viewers, that means more direct control over content discovery; for creators, it adds a new, prompt-driven layer that could influence who gets surfaced to different audiences.
How to Create and Pin a YouTube Custom Feed
To start using the AI feed builder, you need to be signed in and have your YouTube watch and search history turned on in your account settings. Then, on the Home page, tap the “Your custom feed” chip at the top. You’ll see a prompt box plus suggested prompts underneath. You can pick one of these or write your own description, such as “15-minute HIIT workouts that don’t need any equipment and zero jumping” or “deep-dive tech podcasts to learn more about using AI for work.” After you submit the prompt, YouTube generates a custom feed that behaves like a dynamic playlist of recommendations. According to Google, you can maintain one custom feed at a time and “pin it as a saved chip right to the top of your Home page” for quick access over the next 30 days.
Refining Prompts for Better Personalized Playlists
The quality of your YouTube custom feed depends on how clearly you describe what you want. Think in terms of topic, format, and constraints. For example, “help me unwind after work with guided meditations under 10 minutes” tells the AI both the mood and the ideal video length. Adding details like “beginner-friendly,” “no equipment,” “longform interviews,” or “beyond my usual feed” gives the system more to work with and can help you escape your existing viewing bubble. You can adjust the text at any time via the box at the top of the feed, so treat it as an ongoing conversation instead of a one-shot command. If the results feel off, tweak one element at a time—change duration, difficulty level, or subject depth—then watch and interact with a few videos so the system can recalibrate around your updated preferences.
Managing Expiry, Switching Feeds, and Giving Feedback
Your AI-generated custom feed is temporary: it lives for 30 days, after which both the prompt and its feed expire and need to be recreated if you want to keep that viewing pattern. During those 30 days, you can freely switch between the AI feed and your regular Home feed. When you’re inside the custom feed, clicking the standard Home icon takes you back to your normal recommendations. The feed will refresh over time with new videos that match your prompt, rather than staying as a fixed playlist. If the content feels irrelevant or low quality, open the three-dot menu next to the prompt box and pick the feedback option like “Something wrong?” to report issues. Updating the prompt and using standard tools such as “Not interested” or “Don’t recommend channel” will also help steer future results.
What AI Feed Customization Means for Creators and Discovery
AI-powered feed customization adds a new layer to how viewers find videos and how creators get discovered. Instead of relying solely on watch history and generic algorithmic signals, prompts let viewers describe highly specific needs—quick tutorials, niche hobbies, or “something different beyond my usual feed.” This could give smaller or less typical channels more chances to appear when they match detailed requests, especially around emerging topics like using AI for work. At the same time, prompt-driven curation might narrow viewing patterns if people always request similar vibes or topics. Platforms have already faced scrutiny for recommendation systems that reinforce existing interests, and prompt-based feeds could deepen that effect when users ask for very targeted content every day. For creators, that means optimizing titles, descriptions, and topics so they align with the natural language phrases viewers are likely to type into the AI feed builder.