PureFeed vs "Not interested" & algorithm training

Why "Not interested" never seems to work

Every platform offers feedback buttons: Not interested, Show fewer, Don't recommend channel. The theory is that you train the recommender toward the feed you want.

In practice you're negotiating with a system whose objective is engagement, not your wellbeing — and outrage engages. The recommender may honor your click about that video while continuing to serve the pattern, because the pattern performs. Your feedback is one signal among thousands, and it's outvoted by your own worst impulses: every rage-click you make argues against every Not-interested you file.

Where the other approach wins: Feedback buttons do move the needle on clear-cut topical signals — mark ten woodworking videos Not interested and the woodworking mostly stops. Where they fail is tone: there is no button for "less bait, same topics", because bait is what the objective function is optimizing for. Use the buttons for topics; they cost nothing.

	"Not interested" & algorithm training	PureFeed
Who decides	The platform's recommender, weighing your click against engagement	You — thresholds and categories are enforced in your browser
Enforcement	A suggestion the algorithm may ignore	A rule: score outside your limits → post hidden, every time
Time to effect	Weeks of consistent clicking, results drift back	Immediate, from the first page load
Tone vs topic	Topical only — no button for "less rage-bait"	Tone is the whole point: negativity and sensationalism are scored directly
Transparency	Opaque; you can't see what your clicks did	Every hidden post shows its scores and the reason

Training the algorithm is asking the house to deal you better cards. Filtering is playing with your own deck. Keep clicking Not interested on topics you're done with — and put a browser-side filter between you and everything the recommender serves anyway.

PureFeed scores every post for negativity, clickbait, and value on Reddit, YouTube, X, and LinkedIn, and hides what fails your thresholds. Free, reversible, and it shows its work.

Add PureFeed — free →

Common questions

Does PureFeed change what the algorithm sends me?

No — it filters what renders in your browser. The recommender still serves the same list; you just don't have to see the junk. Many users report their habits change anyway once the bait stops landing.

Is this the same as the platform's own quality filters?

Platform filters optimize for the platform's definition of quality, which includes engagement. PureFeed optimizes for yours: you set the thresholds, and the scoring runs on every post regardless of how well it performs.

More comparisons: Keyword blockers & muted words · Site blockers & app timers