The Rise of Platform Clutter
The ubiquity of AI-generated content has fundamentally changed the social media landscape, and LinkedIn, in particular, has become a hotbed for repetitive, hollow, and low-quality posts. Research suggests that a significant percentage of long-form content on the platform is either fully generated by AI or heavily assisted by it, leading to a cluttered user experience. In response, independent developer and consultant Tom Frazier has launched Slop Mop, an open-source Chrome extension designed to bring order back to professional feeds by filtering out what is colloquially referred to as "slop."
How It Works: Beyond Basic AI Detection
Slop Mop does not merely hunt for AI-generated text; it functions as an intelligent filter for content quality. The extension utilizes a sophisticated checklist of twelve criteria derived from SEO research by Graphite Growth. These "tells" include identifying common engagement-baiting tactics, such as the use of artificial "see more" breaks, hyperbolic language, empty promises, and suspiciously neat life lessons. By analyzing these linguistic markers, the tool assigns a "slop score" to posts in real-time.
To ensure high-quality, human-centric content remains visible, the model also performs a sanity check. It evaluates whether a post sounds like it was written by a unique individual and whether it provides genuine utility to the reader. These factors lower the overall slop score, ensuring that authentic, insightful content is preserved. The extension is designed to flag or hide offending posts, leaving the final decision up to the user while maintaining compatibility with LinkedIn’s native "Seems Like AI Slop" reporting feature.
The Power of Jev Architecture
At the heart of the extension lies Jev, a new probabilistic decision-making AI model from TypeSafe. Unlike standard large language models (LLMs) that act as chatbots, Jev is built to make precise, context-aware decisions for software systems via an API. The project serves as a practical demonstration of how this new class of "non-chatty" AI can handle complex, logic-based tasks more efficiently and cheaply than traditional conversational agents.
Frazier has open-sourced the project on GitHub, allowing developers to audit the code and understand how to integrate Jev into their own workflows. Because the extension runs on Frazier’s personal API key, he has implemented a usage cap of 250 posts per day per installation to keep operational costs manageable. Currently, the cost remains impressively low—approximately $0.005 per user per day—making it a sustainable experiment in AI-driven content moderation.
Why It Matters
- Platform Integrity: By prioritizing value over engagement tactics, tools like Slop Mop help re-establish meaningful professional discourse.
- AI Utility: It showcases a shift away from conversational AI toward "AI infrastructure" designed for specific, high-stakes decision-making.
- Platform Agnostic Potential: While currently limited to LinkedIn, the model's success suggests a potential expansion to other social platforms suffering from content decay.
Ultimately, Slop Mop is not just a cleaning tool; it is a proof-of-concept for the future of AI infrastructure. As the internet becomes increasingly saturated with automated text, the ability for software to discern "worth your time" from "digital noise" will become an essential component of the browsing experience.
