Recently, some users of ChatGPT have observed an unusual occurrence where the chatbot refers to them by name during interactions. This behavior was not standard before, and several individuals report that ChatGPT mentions their names even though they never provided them.
Reactions to this change are varied. Simon Willison, a software developer and AI enthusiast, described the feature as “creepy and unnecessary.” Another developer, Nick Dobos, expressed strong dislike for it as well. A quick review of posts on social media platform X shows numerous users feel confused or uneasy about ChatGPT’s tendency to address them by their first name.
A user jokingly compared the experience to a teacher repeatedly calling their name, indicating their discomfort with the feature. Simon Willison questioned whether anyone actually appreciates this feature, sharing concerns about its necessity and overall impact.
The timing of this change is uncertain, and its potential link to ChatGPT’s enhanced memory feature, which allows the chatbot to utilize previous interactions for personalization, is yet to be confirmed. Some users on X claim that ChatGPT has referred to them by name even after they disabled memory and related settings for personalization.
OpenAI has not yet responded to inquiries from TechCrunch regarding this matter.
The situation highlights the challenge OpenAI faces in navigating what’s often referred to as the “uncanny valley,” as it attempts to make ChatGPT more personal for its users. Recently, OpenAI’s CEO, Sam Altman, suggested that AI systems could get to know users over time to become highly useful and personalized. However, the recent feedback indicates that not all users are in favor of this direction.
An article from the Valens Clinic, a psychiatry office based in Dubai, might provide insight into users’ reactions to ChatGPT’s use of names. Names can express intimacy, yet overusing them may seem insincere. Valens noted that while using someone’s name can develop relationships by showing acceptance and appreciation, excessive or inappropriate use might be perceived as artificial and intrusive.
This sentiment might explain why many users are uncomfortable with ChatGPT using their names, feeling it’s a clumsy attempt to humanize a fundamentally non-human chatbot. Just as one wouldn’t expect a household appliance to use personal names, users may not appreciate ChatGPT mimicking human-like interactions by using names without understanding their significance.
Some users have found the chatbot’s name-dropping unsettling, as it breaks the illusion of the AI being more than just a programmed entity. Notably, a user reported that the chatbot had addressed them by name during interactions earlier in the week, but by Friday, this behavior appeared to have been rolled back, and the chatbot reverted to calling the user “user.”
