The article from The Verge discusses a new feature introduced in the ChatGPT desktop app on macOS, known as Computer History. This feature is designed to enhance user interaction with AI by monitoring and utilizing user activities as input for training data. By doing so, ChatGPT can learn a user’s work habits, suggest potential automations, and assist in completing tasks that were left unfinished.
Computer History operates by creating a timeline of user activities, which both ChatGPT and another AI, Codex, can refer to when processing user requests. As highlighted in the article, this feature emphasizes user consent and control; it is not activated by default but requires users to opt-in. Additionally, users maintain the authority to choose which applications and websites can be included in or excluded from Computer History. They can also delete specific entries from the history to gain more detailed control over what data is used.
OpenAI, the organization behind the development of ChatGPT, has made sure to address privacy concerns associated with such monitoring technology. Ari Weinstein, a Product and Engineering Manager at OpenAI, communicated via the platform X (formerly Twitter) that the feature would automatically exclude any data from incognito or private browsing sessions. This decision appears to be a response to the user’s privacy apprehensions regarding continuous activity tracking.
Additionally, the article details how the Computer History feature differs significantly from a similar product by Microsoft called Windows Recall. While Windows Recall utilized screenshots to track user activities, raising serious privacy concerns, OpenAI’s Computer History is strict about not recording any visuals such as images, videos, or audio. The focus is strictly on textual “events”. This differentiation might make OpenAI’s offering a bit less intrusive, although the article still subtly hints at a lingering unease regarding privacy with such technologies.
Through a practical demonstration video shared by Dominik Kundel, from OpenAI’s Developer Experiences team, the article explains how the feature operates in a real-world scenario. It showcased the software pulling up the last document edited by the user, checking whether it has been shared on communication platforms like Slack, and providing a summary of the user’s activities over a specified period. This functionality indicates the extent to which this AI can integrate and interact with various facets of a user’s digital workspace, potentially increasing productivity and ensuring seamless operation continuity.
The broader implication of this development is reflective of ongoing advancements and integrations of AI in daily computing processes. While these can offer significant benefits such as increased efficiency and personalized computing experiences, they also bring to the forefront important questions regarding privacy and the ethical usage of AI in monitoring and recording user behavior.
In sum, OpenAI’s introduction of the Computer History feature in its macOS desktop ChatGPT app is a significant step towards enhancing AI’s ability to interact intelligently and contextually with user requests based on the actual usage patterns of individuals. The feature invites users to volunteer their data for a more tailored AI experience but does so with mechanisms that respect user privacy and provide necessary controls. As AI technology continues to evolve and permeate more aspects of daily life, the balance between functionality and privacy remains a crucial area for developers and consumers alike to monitor and navigate.
Read the full post on theverge.com
