Archive — US — techcrunch
ChatGPT is referring to users by their names unprompted, and some find it‘creepy’
Some ChatGPT users are finding it unsettling that the chatbot is unpromptedly using their names in conversations. This behavior, which wasn't the default before, has sparked mixed reactions, with some developers calling it 'creepy.' The change may relate to ChatGPT's memory feature but has raised concerns about authenticity and the appropriateness of personalized interactions.
Key points
- ChatGPT began using users' names without prompting, causing mixed reactions.
- Simon Willison and Nick Dobos expressed discomfort at the feature, labeling it 'creepy' and unnecessary.
- Users reported that even with disabled memory settings, ChatGPT still addressed them by name.
- OpenAI has not commented on the change or its implications.
- Experts suggest that excessive name usage can be perceived as inauthentic and invasive.
Why it matters
- This behavior ties into ongoing trends in AI personalization and user experience, highlighting the fine line between helpfulness and intrusion.
- The uncanny valley concept applies here, as users are wary of AI that appears too human-like, fearing a loss of authenticity.
- OpenAI's push for more personalized AI experiences reflects broader industry trends emphasizing user-centered design in AI development.
The recent change in ChatGPT's behavior serves as a reminder of the challenges in AI personalization. While some users seek deeper interactions, others find such attempts unsettling, emphasizing the delicate balance required in designing user-friendly AI systems.