DARLING DEAR
MY HEART YEARNS FOR YOUR THIRST. MY IMPATIENT EAGERNESS PASSIONATELY YEARNS FOR YOUR APPETITE. MY DESIRE ATTRACTS YOUR SYMPATHETIC LONGING. YOU ARE MY SEDUCTIVE SYMPATHY. MY PRECIOUS WISH.
Advertisement
YOURS CURIOUSLY
MUC
This was one of a number of enigmatic notes pinned to the computing department noticeboard at the University of Manchester, UK, back in August 1953.
There was no great mystery about āMUCā, however: that could only be āManchester University Computerā, the worldās first commercial programmable electronic computer. Designed to work on atomic bombs, X-ray crystallography and other serious science, what business had this Ferranti Mark 1 writing love letters?
The answer, of course, was a gifted, under-occupied programmer. It was both perverse and predictable that just five years after the birth of the modern computer, someone would use artificial intelligence to produce what machines neither need nor want.
That tricky question of the relationship between the human and the artificial is key to two new exhibitions in Manchester, this yearās European City of Science.
The Imitation Game takes its name from an experiment proposed by Alan Turing in 1950, while he was pioneering artificial intelligence research at the University of Manchester. Better known as the Turing test, posed the challenge of getting a machine to converse with a human so naturally that the exchange could be mistaken for one between humans.
One way or another, the artworks all play the game, but none of them take it too seriously. āTheyāre all totally flawed!ā laughed curator Clare Gannaway. āItās just as much about what we are able to project on to those things.ā
Synthetic valentines like the one above appear in David Linkās LoveLetters_1.0. His installation includes a recreation of the bulletin board used for programmer Christopher Stracheyās enigmatic notes, along with the original teleprinter that output the messages, and an array of 1950s-style cathode ray tubes suspended in mid-air to display the program as it ran on the Ferranti Mark 1.
Vintage tech lovers will linger, perhaps wistfully considering Manchesterās moment as a postwar Silicon Valley. Elsewhere, thereās plenty on show for fans of robot aesthetics: ās insectoid machines succeed in being both unmistakably biomorphic and industrial, while a boxy roaming robot built by Paul Granjon and autonomous wheelchairs by Mari Velonaki play on our effortless talent for responding emotionally to lifeless objects.
More superficially likelike are the skeletal androids of ās Talk. Kjellmark was inspired by meeting scientists from the international Human Brain Project, which aims to create silicon-based simulations of working brains. As part of this hugely ambitious programme, a University of Manchester team led by Steve Furber built SpiNNaker, a massively parallel computer, to mimic the neural networks between our ears.
This technology animates Talkās robot protagonists as they sit in a cosy living room gesticulating while discussing dreams, brains, consciousness and identity. They are also playing the imitation game: invade their personal space and one turns its head and video-projected eyes to you, before resuming its conversation with a āWhere were we?ā
Talk was too clunky to hold my attention for long, but thereās no doubting the scientists are playing the imitation game for real. In the showās catalogue, Furber lays out the utilitarian reasons for building machines that play human, echoing Turingās motivation to find out if machines can think ā or at least, appear to think. In 2016, thatās not enough for Furberās team, which wants to reproduce real human thought in the detail of its neurological processes.
Only one work fully engages with machine thinking: Lynn Hershman Leesonās veteran . However, despite Rubyās 15 years of AI learning, she remains frustratingly obtuse in conversation. That is a pity: newer chatbots have won the imitation game, and might have engaged visitors more with the artistās conceptual concerns.
But in the end, it doesnāt really matter: this playful show is not meant to advance AI research or to amaze with convincing automata. Instead, it turns the spotlight on the imitation gameās human judges. Wandering through the gallery, the judges are us, and what is being tested is our response to the machines.
Those wanting more shock and awe can find it a short walk away at the Home arts centre, where artist duo Al and Al (Al Holmes and Al Taylor) have painted a suite of rooms black, the better to disorient visitors while they seduce them with phantasmagoria and high production values.
This is Incidents of Travel in the Multiverse, featuring everything from equations displayed in neon to CGI cyborgs declaring their love over breakfast in 2154 and Alan Turing (him again) having a paranoid breakdown in 1954.
Al and Al have also talked to eminent scientists to inform their work: in their case, superstring theorist Brian Greene at Columbia University, New York, and the DNA-scanning physicist Bart Hoogenboom of University College London. No doubt there is plenty of science woven into this intoxicating installation, but visitors may be happy to revel in the immersive kaleidoscope of images, sounds and concepts and not think so hard.
Back at The Imitation Game, the sense of wonder the artists seek to evoke is quieter, more reflexive. āThe exhibition draws attention to our amazing qualities,ā says curator Gannaway, āto be able to imagine what these funny-looking objects might be thinking or feeling, or what theyāre saying to us.ā
Out in the world, AIs are not playing the imitation game. Instead, they anticipate our every move, purchase and political sympathy. Art reminds us of the clever things only we can do ā for now.
[exhibition_info title=āThe Imitation Gameā gallery=āManchester Art Galleryā gallery_link=āhttp://manchesterartgallery.org/ā location=āManchester, UKā fromdate=ānowā todate=ā5 Juneā]
[exhibition_info title=āIncidents of Travel in the Multiverseā gallery=āHomeā gallery_link=āhttp://homemcr.org/ā location=āManchester, UKā fromdate=ānowā todate=ā10 Aprilā]
