Mattias Rost

Mattias Rost

Large language models (LLMs) are changing how we interact with computers. As they become capable of generating software dynamically, they invite a fundamental rethinking of the computer's role in human activity. In this conceptual paper, we introduce LLM-mediated computing: a paradigm in which interaction is no longer structured around fixed applications, but emerges in real-time through human intent and LLM interpretation. We make three contributions: (1) we articulate a new interaction metaphor of reflective conversation to guide future design, (2) we use the lens of postphenomenology to understand the human-LLM-computer relation, and (3) we propose a new mode of computing based on co-disclosure, in which the computer is constituted in use. Together, they define a new mode of computing, provide a lens to analyze it, and offer a metaphor to design with.

Oana Andrei, Muffy Calder, Matthew Chalmers et al.

This paper sets out a process of app analysis intended to support understanding of use but also redesign. From usage logs we infer activity patterns - Markov models - and employ probabilistic formal analysis to ask questions about the use of the app. The core of this paper's contribution is a bridging of stochastic and formal modelling, but we also describe the work to make that analytic core utile within a design team. We illustrate our work via a case study of a mobile app presenting analytic findings and discussing how they are feeding into redesign. We had posited that two activity patterns indicated two separable sets of users, each of which might benefit from a differently tailored app version, but our subsequent analysis detailed users' interleaving of activity patterns over time - evidence speaking more in favour of redesign that supports each pattern in an integrated way. We uncover patterns consisting of brief glances at particular data and recommend them as possible candidates for new design work on widget extensions: small displays available while users use other apps.