The AI: More than Human exhibition at London’s Barbican Centre throws its visitors in at the deep end in terms of ideas of artificial intelligence or perhaps non-human intelligence.
The first thing visitors are introduced to is the Golem, a folkloric being fashioned from mud or clay that is brought to life to be a protector of Jewish people.
The Thing from the Fantastic Four and a monster from Minecraft were displayed as examples of Golems. This had me thinking: Is there a binary notion of human versus artificial intelligence? Can intelligence exist in inanimate objects because we believe it to?
The best thing for me about the exhibition was the timeline, detailing the recent history of artificial intelligence from the publication of Mary Shelley’s Frankenstein to the present day. It was presented on a touchscreen as well as on a series of adjacent monitors. It would have been brilliant if attendees could take home a paper version or be able to access a digital version after the event.
There were so many great moments in the history of AI included that it is a shame that I was only able to take in and remember a couple. One that truly stood out though was the origin of the term ‘debug’. It came about when Rear Admiral Grace Hopper was operating the Mark I – which punched holes into paper tape - suddenly stopped working. She found that a moth had flown into the machine and its body was blocking one of the holes. There was an actual bug in the system.
Another fascinating and interactive installation was all about word association and allowed visitors to input words into a registry and see where that word sat in terms of four binary spectrums. Those were ‘expensive-cheap’, ‘bad-good’, ‘he-she’, and ‘machine-human’.
I typed in ‘data’ which turned out to be more expensive than cheap, more bad than good, more female than male and more human than machine. I was also surprised to see that ‘magic’ was more machine than human.
Attendees had the opportunity to be the subject of computer-generated art and to create AI art themselves with wire and dust cloths. Many people, including me, chose to drive on a simulated race track and simultaneously have their facial expressions analysed by an AI. There were numerous robots on display including Atila, an experimental planetary explorer, SoFi – a soft robotic fish, and a robot puppy.
Definitions of AI-related terms were dotted around the walls. I would love to know what a data professional thoughts of those definitions in terms of accuracy and complexity, considering they are aimed at the general public.
Ethics, negative bias and morality were also covered in this exhibition with a stand about Joy Buolamwini, the black coder who had to wear a white mask for facial recognition to work on her, and a paragraph about autonomous weapons.
All in all, what was most heartening was the sheer number of people in attendance, especially children, wanting to increase their knowledge and understanding of AI. I got speaking to a retired engineer who is very excited about the possible benefits that AI-powered technologies, such as autonomous vehicles, could bring. She said she often discusses these possibilities and the ethics of them with her family.
The discussions taking place in the data community are also taking place around dinner tables and exhibitions such as these that increase comprehension of a complex but necessary subject matter can only be a good thing.