The suspension of the will to know?
A video of a talk on algorithmic thinking
Last week I gave an online talk as part of a seminar series on ‘Technique and Society’ at the Escuela de Gobierno at the Universidad Adolfo Ibáñez. The seminar series runs throughout 2022 - full details of the series and upcoming speakers can be found here. My talk was on ‘Algorithmic thinking, neural networks and the presences of nonknowledge’.
The session, including both the talk and a detailed discussion, was recorded and is now available on Youtube (available here or below).
The talk is based around a chapter written for my forthcoming book on algorithmic thinking. From memory, the idea for this particular piece originally came about after reading a long article in the Financial Times on neural networks. A bit of a chance encounter. I don’t mention this in the talk, but that is where the idea grew from over about 3 years (it is a source I use in the written version). There was a small comment in that long-read about how these systems cannot be fully explained (not in terms of a black-box, I wouldn’t say, but more in terms of the depths of advancingdeep learning etc). It struck me as interesting that such powerful forms of machine learning might not just be producing knowledge but might also produce tensions around what becomes unknowable. That then became one of the tensions I develop in the book.
Georg Simmel's long essay on secrecy was also a bit of a turning point. He talks in there about the way that confidence is established in the balance between what is known and what is not known. That balance changes in different times, he points out. That observation provided me with a way into the problem.
The talk finishes by arguing that advancing automation brings with it a kind of suspension of what Foucault once called the ‘will to know’. I argue that the desire to know is usurped by a desire to automate. I explain this more in the talk and in the book. It's something I’m hoping I might get chance to work on more in the future, depending on how things go. I feel like it needs more sustained attention to really develop it fully. I need to work out how I can make that happen, otherwise it might become another loose thread.