Lucy Suchman

In the AI Story

Suchman's critique of the AI paradigm came from a distinctive position: she was an anthropologist working inside one of the world's leading technology research centers, observing the actual practices through which human-machine systems were designed, deployed, and used. This gave her critique an authority that purely external criticism could not have: she was not theorizing about the failures of the AI paradigm from outside. She was documenting them as they occurred, in the specific systems being built by specific researchers at one of the field's most celebrated institutions.

Her 2007 update, Human-Machine Reconfigurations, extended the analysis to contemporary robotics and AI. The central argument persisted: human-machine systems should be evaluated not by whether the machine "passes" some threshold of intelligence but by how the interaction is actually organized — who does what work, who supplies what knowledge, whose understanding of the situation is treated as authoritative, and whose labor becomes invisible in the process. This framework has become increasingly relevant as the AI discourse has shifted from debates about machine consciousness to more operational questions about the social and political arrangements through which AI systems are integrated into work and life.

Her subsequent work has addressed questions of military automation, the politics of "autonomous" systems, and the ways in which AI hype functions as a form of political rhetoric. Her 2023 book Contested Relations extends the situated-action framework to questions of global conflict and the use of AI in warfare, arguing that the fiction of machine autonomy obscures the human decisions and human labor that all AI systems require.

For Lave's framework, Suchman is the scholar whose work most directly connects the anthropological critique of cognitivism to the specific technical and institutional features of contemporary AI. Lave and Suchman knew each other's work well, cited each other with approval, and developed complementary challenges to the same fundamental target: the assumption that intelligence consists in the manipulation of context-free representations. The endurance of their challenges is the reason their frameworks have acquired new urgency in the era of large language models.

Origin

Suchman was born in 1946, earned her PhD at UC Berkeley in 1984, and worked at Xerox PARC from 1979 until 2000. She is Professor Emerita of Anthropology of Science and Technology at Lancaster University.

Key Ideas

Human action is situated. Plans are resources for action, not the structure that produces it. Actual behavior is improvised in response to specific circumstances — a finding Suchman documented ethnographically in 1984 and that remains empirically robust.

Ethnography inside technology development. The anthropologist positioned inside the AI research community can produce evidence about the gap between design assumptions and user behavior that no external critic could obtain.

Human-machine reconfigurations. The relevant question is not whether machines are intelligent but how work, knowledge, and authority are distributed across human-machine systems.

Autonomy is a political fiction. Systems described as autonomous are configured by human designers, deployed by human operators, and dependent on human labor that is typically rendered invisible by the autonomy rhetoric.