The Limits of the Chatbot: Questioning the AI Interface

AI language models, while impressive in their ability to follow complex discussions, face an interface barrier that limits their effectiveness. A user can spend an hour discussing various topics such as design, engineering, and psychology with a language model, which manages to track and connect ideas expressed at different times. These models can even challenge the user or articulate thoughts that are not clearly expressed.

However, this intelligence is confined to a chat interface, resembling an enhanced version of a 2015 chatbot. As the conversation lengthens, the scrollbar shrinks, and valuable ideas risk being lost among numerous messages and repetitive assertions.

This interface problem is not just an ergonomic issue. For experts trained in linguistics, psychology, and communication theory, it is also a matter of effective communication. While language remains the primary interface for interacting with technology, the inefficiency of chatbots could limit the full exploitation of the capabilities of AI language models.