Chat is the least interesting interface to LLMs (www.haskellforall.com)

A recent discussion highlights that while chat interfaces are the most common way to interact with large language models (LLMs), they may not be the most effective. The critique emphasizes the shortcomings of prose as a user interface, such as its cumbersome nature on mobile devices and inherent ambiguity. Users typically prefer structured inputs like buttons and sliders, as well as structured outputs such as graphs and tables, which can enhance clarity and usability. This insight is significant for the AI/ML community as it calls for a shift in how developers approach LLM interactions. By leveraging the capabilities of LLMs to accept structured inputs and generate structured outputs, developers can create interfaces that improve user experience and confidence. This approach not only streamlines the interaction process but also encourages greater user satisfaction and trust, making it essential for the evolution of LLM applications beyond simple chat functionalities.