OpenAI has thrown a curveball into the world of chatbots with the release of a new “memory” feature for ChatGPT. This means the chatbot can now retain information gleaned from past interactions, aiming to create a more seamless user experience by saving you from repeating yourself. Sounds convenient, right?
But hold your horses. This memory feature also raises a red flag for privacy concerns. Just like any technology collecting data, the question arises: what exactly will ChatGPT remember about you, and who gets to access it?
While OpenAI assures users they can manage the memories and even “tell it to forget conversationally,” the vagueness surrounding what constitutes an “AI memory” is unsettling. The company’s own video showcases snippets of personal details like family preferences and even location hints, painting a picture of just how intimate these memories can get.
To ease anxieties, OpenAI offers a “temporary chat” option and claims it won’t proactively remember sensitive data unless explicitly requested. But how reliable is this safeguard? For now, this memory feature is in its testing phase, limited to a select group of users. But if you use ChatGPT, keep an eye out for it and carefully consider whether you want this digital companion to hold onto your personal snippets. The convenience of saved information might come at a cost you’re not comfortable with.