Latency at the interface

On June 5th, I’ll be participating in Latent Futures, an online seminar organised by Professors Jen Ross and Richard Sandford. This event challenges the conventional view of futures as distant projections, exploring instead both processes already unfolding and possibilities concealed within our present arrangements.

In both technological history and our theoretical frameworks, we’ve artificially separated two understandings of latency: as measurable delay (a millisecond lag in telecommunications) or hidden potential (the compressed possibilities in statistical models). My talk reunites these bifurcated genealogies, positioning latency as a “generative threshold condition” that emerges wherever different systems, timescales, and ways of knowing come into contact.

Working at these thresholds through recent collaborative projects – from “Latent Intimacies” at Medialab Matadero to experimental software development with Tim Cowlishaw – I’ve come to see that interfaces aren’t passive boundaries, but active sites where constraints and possibilities are negotiated. Millisecond gaps in video calls, dimensionality compression in image recognition, designed impermanence in experimental systems – these aren’t problems to be solved, but generative conditions that shape what becomes possible.

Drawing on examples from film, AI image captioning, and my own experimental prototyping experiences, I’ll show how latency operates simultaneously as constraint and possibility. This approach offers a way to recognise futures already unfolding at the boundaries of perception and practice – not by projecting possibilities onto an empty future, but by cultivating attention to threshold phenomena in the present.

‘Latency at the Interface: Constraint, Possibility, and Perception’

This paper rethinks latency – not just as delay or hidden potential, but as a quality of interfaces where different systems, timescales, and ways of knowing meet. Across its history, “latency” has described both measurable delay (as in experimental psychology, telecommunications) and hidden or compressed potential (in psychoanalysis, statistics, and, more recently, generative AI). Rather than treating these technical and metaphorical genealogies as separate, I argue that they share a common logic. Latency marks a threshold: it is at the interface – where constraint and possibility are negotiated – that what is imperceptible can, under certain conditions, become perceptible.

Drawing from my own work using intentional delay and decay to surface alternative data practices, across collaborative AI prototyping and experimental software, I show how this threshold logic can ground the analysis of latent futures already underway. This perspective clarifies the stakes of working with latent futures, offering a portable analytic for researchers seeking to recognise and cultivate possibilities at the boundaries of systems, disciplines, and temporal regimes.

My analysis positions latency as simultaneously constraint and possibility. This understanding emerged through hands-on, necessarily partial engagements with latent phenomena. Established approaches failed to explain what was observed, resulting in what Helen Verran terms ‘disconcertment’: moments of unease that signal the limits of existing frameworks. Here, partiality becomes a resource for surfacing threshold phenomena. Latency, then, is not a delay to be eliminated, but a generative condition.

Building on John Law’s ‘method assemblage’ and Sara Ahmed’s analysis of orientation, I argue that research approaches themselves operate as interfaces, amplifying certain patterns while rendering others less discernible. I close by outlining how cultivating interface conditions and attending to constraints can help discern possibilities that would otherwise remain imperceptible, inviting engagement with latent futures already unfolding at the boundaries of perception and practice.

References