ChatGPT's New Feature Tracks User Activity for AI Training

ChatGPT's desktop application for macOS has introduced a new "Computer History" feature. This functionality records user clicks and keystrokes, transforming this activity into training data for the AI. The goal is to enable ChatGPT and Codex to learn individual work patterns, suggest automations, and even complete tasks that users have left unfinished by referencing a generated timeline of their actions.

This development raises questions about data privacy and the extent to which AI systems are integrated into daily computing workflows. While the feature aims to enhance user productivity and personalize AI assistance, its data-gathering capabilities mark a significant step in the evolution of AI interaction with personal computing environments. The integration of such detailed activity tracking into a widely used AI tool highlights ongoing trends in machine learning and user experience design.