Remove AI Slop From LinkedIn Using Browser Extension

A Chrome browser extension named Slop Mop has been created by AI advocate Tom Frazier to detect and label subpar content on LinkedIn. The extension uses Jev, a decision-making AI model, to flag posts that exhibit low-quality writing traits. Rather than attempting to discern whether a piece was written by AI, Slop Mop looks for patterns indicative of subpar content.

The extension's criteria include excessive hype, engagement bait, empty praise, overly polished lessons, flowery language, stiff wording, contrast-based definitions, and posts intended to be scrolled through. Slop Mop employs additional inquiries to ascertain if the content resembles a particular individual's writing style and whether it provides value to readers.

The extension's scoring system can flag or hide posts if they surpass the threshold, yet these posts remain on LinkedIn and can still be reported via LinkedIn's "Seems Like AI Slop" feature. Frazier has made the project's code open-source on GitHub to enable developers to analyze and experiment with Jev. Currently, Slop Mop functions using Frazier's personal Jev account and API key, allowing a maximum of 250 analyzed LinkedIn posts per day.

Despite the free service's daily cost of approximately $0.005 per user, Frazier may reconsider the free offering if the extension gains widespread popularity. For now, the project primarily serves as a means for Frazier to investigate Jev's settings and construct a dataset.