The System Model and the User Model: Exploring AI Dashboard Design

TL;DR

This paper advocates for AI dashboards displaying the System and User Models to improve interpretability, usability, and safety in dialogue-based AI systems, emphasizing the importance of interface design beyond text interactions.

Contribution

It introduces the concept of System and User Models as key components for AI dashboards, proposing their display as a novel approach to enhance AI interpretability and safety.

Findings

Dashboards should display System and User Models for better AI transparency.

Interpretable models of AI systems can improve usability and safety.

Designing interfaces around these models is a crucial research direction.

Abstract

This is a speculative essay on interface design and artificial intelligence. Recently there has been a surge of attention to chatbots based on large language models, including widely reported unsavory interactions. We contend that part of the problem is that text is not all you need: sophisticated AI systems should have dashboards, just like all other complicated devices. Assuming the hypothesis that AI systems based on neural networks will contain interpretable models of aspects of the world around them, we discuss what data such dashboards might display. We conjecture that, for many systems, the two most important models will be of the user and of the system itself. We call these the System Model and User Model. We argue that, for usability and safety, interfaces to dialogue-based AI systems should have a parallel display based on the state of the System Model and the User Model. Finding ways to identify, interpret, and display these two models should be a core part of interface research for AI.