do you guys also lowkey feel angry after finding out a post was ai generated, or someone was doing a stupid promotion in the name of a genuine story, or a post is absolutely brainrot wasting your time?
Just like the one i shared in the image above. It somehow looks genuie right? but you and others can vote it out as AI Slop and lowkey why not, because these karma farmers are eating up our feed and our algorithm through engagement bait, and there's nothing stopping them.
i know we are not productive 24/7, but that's exactly why it hurts more. the one time you open reddit or X to just chill, you get fed this stuff.
so i'm almost done building a community powered tool (no AI) to tackle this thingy. basically a browser extension, that will subtly label a post on reddit and X before you even open it. the voting on the label quality is done by genuine people like you and me who have already seen that post.
- the labels that show up on posts can be: "Deceptive Ad" or "Rage Bait" or "AI Brainrot" or "Genuinely Useful" and 130+ different labels depending on community votes
- the voting categories (3 max per user per post): Genuine, Helpful, Wholesome, Ad & Promo, AI Slop, Bait, Brainrot, Misleading, Rant
covers almost every type of post on reddit and X i think, and small enough options that it's literally just 2 button clicks.
Some more stuff though:
- lifetime free for users, but will be rate limited or have a daily vote limit
- THE BROWSER EXTENSION WILL BE OPENSOURCE, the backend will be closed source
- the only data i take is: reddit/X post ID (no content, only url ID), votes (upto 3), ranged upvote count (if 67 i'll store it as 50-100), ranged time, backend auth creds and time spent on post before voting. THATS IT. ain't breaking privacy or TOS of reddit
- the api, server and data will be closed source and used for commercial purposes, like 30 dollars per 1000 posts for businesses IDK haven't fully planned yet but i have a fair reason for why the paid plan makes sense.
- can promise minimal network and visual load
- the vote-to-label algorithm is secure and foolproof, even if wrong people spam "genuine" on an ai slop post, my backend wouldn't get tricked by it
An average user flow:
user opens reddit/X feed > scrolls as usual > about to open a post > reads the label on it > "AI Promo" > skips > opens another post, reads it, feels genuinely useful > votes genuine+helpful and moves on
that's it. 0 friction once installed from the chrome web store.
need your honest views annd would you actually use something like this daily? do you even like it? I promise no problems once past the installation process, high quality feed on Reddit and X always, and you'll end up saving on MENTAL PEACE and TIME heheheh