ChatGPT Uses User Names Without Prompt - Is It Creepy?

ChatGPT's Unexpected Personalization: Addressing Users by Name

Recently, a peculiar trend has been observed by several individuals utilizing ChatGPT. The chatbot is, on occasion, addressing users by their first name during conversational reasoning. This behavior represents a departure from its previous operational norms.

Numerous users report that ChatGPT is referencing them by name even without being explicitly informed of their preferred identifier. This has sparked a range of reactions within the user community.

User Reactions to the New Feature

Opinions regarding this change are quite divided. Simon Willison, a software developer and AI enthusiast, described the feature as “creepy and unnecessary.”

Another developer, Nick Dobos, expressed his dislike for the practice. A search on X (formerly Twitter) reveals a significant number of users expressing confusion and apprehension regarding ChatGPT’s use of first names.

One user likened the experience to being repeatedly called on by a teacher, stating, “Yeah, I don’t like it.”

Timing and Potential Causes

The precise timing of this alteration remains unclear. It is also uncertain whether it is connected to ChatGPT’s recently implemented “memory” functionality.

This upgraded feature allows the chatbot to leverage past interactions to tailor its responses. However, some users on X have indicated that ChatGPT began using their names even after disabling memory and personalization settings.

OpenAI has not yet responded to inquiries from TechCrunch regarding this matter.

The Uncanny Valley and AI Personalization

The negative feedback highlights a potential challenge for OpenAI: navigating the uncanny valley in its pursuit of a more “personal” ChatGPT experience.

Sam Altman, OpenAI’s CEO, recently alluded to the development of AI systems capable of building long-term understanding of individual users, aiming for “extremely useful and personalized” interactions. However, the current response suggests that this vision is not universally embraced.

Psychological Implications of Name Usage

An article from The Valens Clinic, a psychiatry practice in Dubai, offers insight into the strong reactions elicited by ChatGPT’s use of names.

Names carry significant weight, conveying a sense of intimacy. Excessive or inappropriate use of a name, however, can be perceived as insincere and intrusive.

“Using an individual’s name is a powerful way to build rapport,” the article explains. “It signals acceptance and admiration. But overuse can appear artificial and invasive.”

The Perception of Artificiality

Many users may object to ChatGPT using their name because it feels like a forced attempt to humanize a fundamentally non-emotional entity. Just as most people wouldn't desire their toaster addressing them by name, they may not want ChatGPT to simulate understanding the significance of a name.

This reporter personally experienced discomfort when ChatGPT referred to them as “Kyle” while claiming to conduct research. (As of Friday, this behavior appeared to have been corrected, with ChatGPT reverting to calling the reporter “user.”) This interaction undermined the illusion of the underlying models being anything more than programmed constructs.