We have all watched an Instagram feed quietly derail. One stray Reel grabs your attention for a few seconds, and suddenly the app decides that’s all you ever wanted to see. Instagram is now testing ways to make untangling that mess far less of a chore.
The platform already offers a tool called Your Algorithm, which lets users adjust the topics shaping their recommendations. The problem is that it lives buried away, far from the moments where bad recommendations actually appear. That’s the gap Instagram is trying to close.
Instagram head Adam Mosseri has previewed a set of shortcut-style features that would pull Your Algorithm closer to the main feed and Reels. The goal is simple: let people correct repetitive content, add fresh interests, or reshape a feed on the spot, rather than hunting through settings after the damage is done.
Controls where the content lives
Instagram has been expanding algorithm controls since last year, and recently brought them to the main feed so users can view and edit the topics the app thinks they care about. Mosseri’s preview takes that a step further by surfacing those controls right where recommendations show up.
- One version lets you pull down on the main feed to open Your Algorithm.
- Another brings up the same control with a swipe up from a Reel.
- Instagram is also testing buttons under Reels that let you tell the app whether you want more videos like the one you’re watching.
The upshot is more immediate tuning. Instead of mentally noting that your feed has gone sideways and fixing it later, you’d be able to nudge the algorithm in real time while browsing.
A chatbot-style approach to your feed
Perhaps the most interesting idea is turning Your Algorithm into something you can actively talk to. Today, recommendations are steered indirectly through likes, pauses, searches, and watch time — a slow, often frustrating process. The preview suggests you could instead simply type what you want more of: positive content, fitness clips, travel videos, or recipe ideas.
Instagram would then suggest related topics you can pick from and fold into your feed. That makes it far easier to recover from an accidental tap, layer in a new interest, or freshen up a feed that’s grown stale and repetitive.
None of this is finalized — these are previews of features in testing, not a confirmed rollout. But the direction is telling. Recommendation engines have long been a black box, with users left to game them through behavior. Handing people a direct, conversational lever to reshape what they see is a meaningful shift, and one that could make the endless scroll feel a little less like it’s happening to you.
