Research · 03
How does big data move across space and time?
Velocity conventionally describes the speed at which big data are generated and processed. Taken more literally, however, velocity also has direction, placing information in space as well as time. Big data are sharded across storage systems, replicated for availability, cached near anticipated demand, migrated between facilities, and recombined when requested. A record can exist in several locations and states at different moments, making its movement inseparable from where it resides, which version is retrieved, and what a user encounters. Information in Motion takes this broader view of velocity, studying how changing copies and states circulate and remain coordinated across systems separated by distance.
Web 2.0 services are engineered to make much of this complexity disappear at the user interface. Search results and streaming media are expected to respond almost immediately, an experience made possible by extensive computational optimization and the strategic placement of infrastructure. Physical distance nevertheless imposes a lower bound because information cannot travel faster than the speed of light. Real-time systems operate within that constraint by coordinating distributed processes so that differences in timing remain within the tolerances required by applications and their users. BDSL distinguishes physical propagation from network latency, application response time, and perceived delay, examining how replication, caching, prediction, routing, and synchronization manage these different layers. The apparent immediacy of big data is an infrastructural achievement, dependent on extensive computation and carefully situated systems whose work becomes least visible when it succeeds.
In other words, “real time” has a geography as well as a temporality. Data travel along fibre-optic networks and among facilities, corporate networks, and legal jurisdictions that shape where information can reside, which routes it can take, and under what conditions it remains accessible. Systems designed to minimize delay can compress the experience of distance without removing differences in infrastructure, regulation, local time, or physical location. As networked services approach simultaneity, the remaining differences in timing and location acquire greater significance. Information in Motion follows these differences to ask what the approximation of “real time” reveals about the physical and institutional geography of an information environment engineered to make distance appear inconsequential.
Examines bodies of evidence that become available for research by routes never intended to create an archive, from rescued websites and data dumps to leaks, hacks, and legal proceedings. The project uses these unintended afterlives to investigate digital records and information systems otherwise beyond scholarly observation.
Traces the development of personalization from Web 2.0 recommendation and engagement systems to generative AI. The Personalized Web examines how systems have learned to observe, infer, and predict their users, and what changes as behavioural traces are increasingly joined by personal disclosure in sustained interactions with machines.
Investigates how globally distributed information systems approximate “real time” despite the physical limits of distance, combining computational experiments and documentary research to trace how network architecture, synchronization, human perception, and jurisdiction shape the experience of a shared present.
Follows the material demands of big data and AI from energy, water, and data centres to semiconductors, critical minerals, and global supply chains. What Computing Takes investigates how large-scale computing reshapes the places and communities that sustain it, from resource use and noise pollution to industrial development, supply security, and geopolitics.