Show HN: HN Bot Detector - Detects LLM-Generated Comments on Hacker News (github.com)

A new tool called the HN Bot Detector has been launched to identify comments generated by large language models (LLMs) on Hacker News. Users can paste a comment URL, ID, or raw text to receive a bot score along with a detailed breakdown of signals that influenced the score. This development arose from the creator's experience of being banned from Hacker News due to LLM-assisted postings, serving as an innovative response to concerns over automated content generation. This tool is significant for the AI/ML community as it pushes the boundaries of detection capabilities for generated text, emphasizing the importance of content authenticity on online platforms. It employs advanced techniques, including character n-gram TF-IDF vectors and heuristic scoring based on common characteristics of LLM outputs, such as formal hedges, buzzwords, and typographic features. Additionally, it allows users to analyze a user's last 50 comments or scan comments on specific posts, thus facilitating greater transparency in discourse and combatting potential misinformation introduced by AI-generated text. The open-source nature of the project, built with Next.js 14 and TypeScript, encourages further research and development in detecting synthetic content online.