TLDR: WASHINGTONâYouTube is rolling out Gemini powered custom video feeds in the U.S., generated from your prompts and refreshed continuously. You can edit prompts, but YouTube Kids gets no new AI labels yet.
Key Takeaways:
- YouTube is testing AI driven personalization against rivals like Reddit, Bluesky, and X style discovery tools.
- Users tap Your custom feed, prompt Gemini with moods or topics, then pin a continuously refreshing feed to the homepage.
- Prompt control and editable vibes shift discovery toward user authorship, while AI labeling upgrades still exclude YouTube Kids.
This is YouTube conceding that ârecommendedâ is no longer enough. With Gemini, you get to steer the algorithm in plain language, then watch it keep rearranging itself.
This is YouTube conceding that ârecommendedâ is no longer enough. With Gemini, you get to steer the algorithm in plain language, then watch it keep rearranging itself.
Q&A
What happens if your prompt keeps drifting after months of watching?
Because the feed refreshes continuously, your ongoing edits and new prompts likely reshape recommendations over time, turning discovery into an evolving conversation rather than a one time search.
Why does YouTube Kids miss out on the new labels even as YouTube upgrades labeling on main?
YouTube says Kids constraints, like different safety policies and limited content descriptions, block the same labeling approach from applying immediately.
Could custom feeds reduce the ârabbit holeâ problem that recommendation systems amplify?
Possibly, since users explicitly define intent, like âguided meditations under 10 minutes.â But refresh and prompt drift could still expand variety if the model interprets your vibe broadly.
How might this change creator strategy compared with standard algorithmic recommendations?
Creators could benefit from prompts that map to recognizable niches and formats, but may also face churn if a feedâs user tuned preferences shift away.
What precedent does this follow from tools like TweetDeck style control and community curation platforms?
It mirrors a broader move from passive discovery to user configured feeds, borrowing the idea of a personalized dashboard while keeping YouTubeâs continuous recommender behavior underneath.
No comments yet. Be the first to share your thoughts!