ANU School of Sociology x ANU Computational Culture Lab seminar on 'Climate’s Computational Infrastructures', Prof Michael Richardson (UNSW) and Prof Adrian Mackenzie (ANU)

[Thao Phan]

Please join us for this ANU School of Sociology x ANU Computational Culture Lab event on 'Climate’s Computational Infrastructures' featuring works in progress from Prof Michael Richardson (UNSW) and Prof Adrian Mackenzie (ANU).

• Date: Thursday 9 April, 2026
• Time: 12 - 1:30pm (AEST)
• Location: RSSS Room 4.69, and Zoom
• Register to receive zoom link: https://events.humanitix.com/anu-school-of-sociology-x-cclab-seminar-climate-s-computational-infrastructures

What roles do computational infrastructures and their embedded values, priorities, and assumptions play in making knowledge about planetary futures? How do computational ensembles pull together computing architectures, data pipelines, calculative practices, and institutional settings to make climate knowledge actionable? What happens to uncertainty and the incomputable richness of ecologies as computational models become decisive in managing Earth futures? 

In this seminar, Michael Richardson and Adrian Mackenzie offer two works-in-progress as invitations for a robust discussion of climate modelling and planetary computation as increasingly operative ensembles for making and managing futures.

---

Computing Earth: what sort of thing is a computable planet?

Michael Richardson (with Anna Munster, Adrian Mackenzie, Baden Pailthorpe, Kathrin Maurer & Antoine Bousquet)

Earth is increasingly staged as a computable object: not a singular replica, but an operational object assembled through technical ensembles that render Earth systems addressable, actionable, and governable. Drawing on Simondon and theories of computability, we argue that planetary computation stabilises prediction as decision by coordinating models, datasets, compute infrastructures, standards, interfaces, and institutional mandates. Using Destination Earth (DestinE) as a diagnostic case, we show how platform gateways, data backbones, and twin engines package probabilistic projections and scenario products for intervention contexts. Incomputability persists as operational remainder, intensified by machine-learning compression, surrogates, and latent spaces that redistribute uncertainty as risk.

Michael Richardson is Professor of Media & Culture at UNSW Sydney and a Chief Investigator with the ARC Centre of Excellence on Automated Decision-Making + Society. His latest book is Nonhuman Witnessing: War, Ecology, and Data after the End of the World (Duke University Press, 2024).

Coupled sensitivities: calculative agencies and overflows in climate overshoot

Adrian Mackenzie 

Climate models function as an experiential membrane. Rather than being mere diagnostic tools, these computational ensembles form a "membrane" that filters and shapes human experiences of time, agency, and mundane feelings of comfort or discomfort. Senses of crisis or futures, and the witnessing of injustice or  loss are now fundamentally inseparable from the calculative agency of models.  Modeling, however,  is a dynamic calculative agency. It uses "framing" to disentangle the climate as  calculable geophysical and socioeconomic processes. This framing inevitably produces "incessant overflows" that can be seen at various points over the last few decades. Anomalies defy existing computational frames and force a recursive re-evaluation of models and the propositions they support. Describing the temporal effects of these overflows, the paper will suggest that 'coupled sensitivity'  reconfigures what counts as experience of climate.  This concept designates a shift where the physical susceptibility of the Earth system to change becomes inextricably linked to the heightened sensitivity and attunement imbued with model projections, scenarios, attributions and pathways. Contemporary engagements with anxiety, inaction, or overshoot might be understood in terms of coupled sensitivity.

Adrian Mackenzie is Professor in Sociology at ANU. His latest book is 1000 Platforms: Ensembles as Ontological Experiences (Bristol University Press, 2025)