Last fall, Vincent deBritto and Ozayr Saloojee invited me to come visit their Resilient Infrastructures project at the University of Minnesota; my main contribution was to deliver the lecture above, “Glitches, Flash Crashes, and Very Bad Futurists”. The lecture examines a particular class of landscape problem, which I’ve provisionally described as “glitches and flash crashes”, argues that this kind of problem reveals a potentially (literally) disastrous flaw in the project of distributed design, and concludes by recommending that architects and landscape architects draw on the tools and methodologies of the discipline of futurism to become better futurists. This brings together and expands on a pair of concepts that I began to develop in earlier posts on mammoth, “very bad futurists” and “unknown unknowns”.

Here’s my full description of the talk:

In the mid-nineteenth century, British geologist Charles Lyell travelled through the antebellum American South; what he found shocked him: forest-clearing, shoddy farming practices, and rainwater had worked together across broad swathes of the South to gouge vast and deep gullies that had no apparent geologic precedent.

This extremely rapid change—it took only a few years for millennia of accumulated soil to completely erode, leaving wounds that gaped often fifty, sixty, even eighty feet deep—might be understood as an example of how landscapes are prone to rapid perversion through the aggregate impact of many actors (in this case, farmers) behaving in accordance with misaligned incentives (in this case, a system of tenantry that favored exhausting plot after plot over developing sustainable practices for a single plot). Collectively, these tendencies both present difficulties for the design of resilient human settlements and suggest the need for designers to become better and more rigorous futurists.