Destabilize AI Seminar Series - Guest speaker Lucy Suchman ​​​​

20-Mar-2025 | 17:00-19:00 CET | Online

Destabilize AI Seminar Series

While LA was battling wildfires and tech billionaires’ wealth broke historical records, one solution remained for solving all our planetary problems: artificial general intelligence. As Trump and his broligarchs embarked on a quest to Mars, traveling through StarGate, they linked the further expansion of AI models and data center development more squarely to fossil fuel production and hegemony, “drill, baby, drill!”. At the AI Action Summit in Paris, European leaders pledged allegiance, enabling the further expansion of AI and computational infrastructures.

Meanwhile, many of us, scholars, policy makers, civil society leaders, and concerned citizens are seeing the cracks in the otherwise mesmerizing new AI functionalities. What is going on here? How should we understand AI and its (geo)politics? What role should AI play in our shared future? And what does that mean for how we work towards alternative visions and practices that truly help us navigate away from deepening planetary crises?

To make meaningful headway in times of turbulence and despair, the Sociotechnical AI Systems Lab kickstarts the Destabilize AI Seminar Series, in which we’ll collectively sit with these large questions, aiming to arrive at meaningful forms of critique and practices for system design, engineering, policy and governance. This series interrupts the rapidly hardening narratives surrounding the promise of AI, with a discussion of what AI really is and what it could be.

Inaugural Seminar: Professor Lucy Suchman

Destabilize AI: Problems, data, and algorithmic solutions

Abstract: This talk is intended as a contribution to a small but growing movement to resist the proposition that ‘AI’ is the driving technology of our age, an assertion that – whether offered as affirmation or alarm – serves the interests of those invested in algorithmic intensification. The point is to tether debate about the efficacy of algorithmic technologies to their politics, raising a set of questions otherwise absent from the discussion. Who defines the problem to be addressed and how is the range of possible solutions predetermined and circumscribed by a priori investments in automation? What are the implications of solutions driven by data, including the political economies of datafication? Given the necessary reductions, and inevitable betrayals, involved in datafication, what does it mean to engage responsibly in the development of algorithmic systems? And what would it mean to put AI in its place, considered as one limited, but potentially useful tool within a wider assemblage of practices for the diagnosis of complex problems and collective work toward their just resolution? Addressing these questions requires engaging with genealogical histories that trace the throughlines of discriminatory social ordering, mechanisation, control, and resistance that give AI its rhetorical/material life.

Location

The event will be a hybrid event. Please indicate on the registration form if you will be joining online or in person.

Online: You will receive a link for the video call following registration.

In person at TU Delft: Join us for drinks and bites at TPM-Lecture Hall A - Henk Sol