As an improbable supercomputer decimates all human opposition in a game of popular quiz-show Jeopardy and solid-state hard drives sap the romance out of rationalising data storage, Tom Lyall is contemplating his own relationship to the machines. Defrag_ is a witty, wise and defiantly intelligent examination of human memory and experience in the age of the social network, where lives are backed up by timelines and diaries trickle out in 140 character chunks.
Lyall has partitioned his show quite firmly. The first 40 minutes are given to a discursive but never rambling account of IBM’s room-sized computer WATSON, of evading essay crises by watching the intricate collation of data blocks on an MS DOS defragmentation screen, and of a brain injury which inspired a sudden distrust of his own memory. The second locates us in science fiction territory, with Lyall, or his nameless narrator, trapped in a cell with a computer that calmly tests his abilities, memories and resources. Like all good techno-speculation, it has a cunning sci-fi plot to hang its smarts on, and the one-two punch of anecdote and action works well.
Lyall links the sections with discussion of a supposed software supercomputer, that allows an individual to back themselves up through the submission of all data relating to their lives. It’s a vision of 21st century electronic storage that sounds eerily familiar. We are subtly reminded that as we teach computers our own stories, they could learn to tell them for us, and to tell their own. Lyall asserts, without pejorative judgement in either direction, that we are programming networks with our own narratives, and that as we share our experiences with one another online, the recording device itself could be listening in, and its ambitions could be greater than the $2,000 Jeopardy-jackpot.
Lyall is a captivating performer, keeping the line between autobiography and fantasy intentionally oblique, and the tone of the show bright and clean, never allowing the human story to vanish beneath the electronics. His discussion of physical disassociation is moving, and the analogues he unearths between human and mechanical computation are consistently persuasive. Stripped down to his pyjamas and kept on a short leash by his GLaDOS like tormentor, the same concepts are re-examined in practise, as we see a conflict of cognitive structures replace the traditional man vs. machine debate.
Ethics are kept firmly and smartly and bay, as Lyall avoids hectoring, easy answers or neat conclusions. There is a sense, by the conclusion, that Defrag_ has not quite turned the screws tight enough, that its own process of sifting and ordering the tenderness, insight and narrative that Lyall keeps suspended is not yet fully complete, but this is clever and sophisticated stuff, that looks forwards to the future without squinting or tipping a wink, with both eyes wide open.
Read the Exeunt interview with Tom Lyall.