Short Abstract

This panel seeks empirical studies on how diverse users of AI practically reason with, evaluate, and manage its messy, unpredictable outputs, moving beyond the hype of 'disruption' to critically assess its real-world limits.

Description

The inflationary claims of ‘disruption’ surrounding the ‘new’ Artificial Intelligence (AI), often propagated by the corporations developing them, hinder our understanding of both the true potentials of these technologies and challenges related to their integration in social life. To get a better grasp on their rapid advancements and their future consequences for various types of jobs, we need to understand how AI is made relevant to its specific contexts of use. While they are publicly presented as seamless or autonomous tools, such technologies are often messy, unpredictable, and prone to generating outputs that users find ambiguous, problematic, or simply incorrect. This creates a critical gap between the AI's imagined or prescribed use and the practical, situated work required for its smooth operation. Attending to the broad sphere of activities that takes place to make AI work can provide a more measured and empirically grounded basis for evaluating its achievements and limitations as part of its entanglements with our everyday lives.

This panel invites empirical investigations that uncover the lived difficulties of working with various applications of AI. We welcome contributions exploring, but not limited to, the following questions:

• How do a wide range of users, from domain experts to laypersons, actually manage and make sense of the results produced by AI technologies in practice?

• What mundane methods and practical reasoning skills do people employ to evaluate, trust, or challenge AI’s outputs?

• How can we empirically study the multiplicity of reasoning styles and ad-hoc procedures users adopt when evaluating AI-produced results?

• What does attending to these practical difficulties reveal about the actual, rather than promised, capabilities of automation and the necessity of situated human competences?

By focusing on the ‘just how’ of AI’s use, this panel will contribute to the ongoing debates surrounding the future of work, human–machine collaboration, and the observable societal implications of ‘disruptive’ technologies.

This Traditional Open Panel has 1 pending paper proposal.
Propose paper