Between Artificial
and Human

Fears and Anxieties

In my work, I focus more often on the fears we have about ‘machines that might think’ rather than on our enthusiasm for them. Fears can include the almost visceral feeling of disquiet or disturbance about beings that fall into a weird space of near human-likeness but not near enough. This unpleasant space is often understood as the ‘Uncanny Valley’, originally identified by Japanese computer scientist Masahiro Mori in the 1970s.

This feeling also appears in and fuels our apocalyptic stories about AI. In a Western context, the most famous, or perhaps infamous given its overuse, an example is the Terminator. The imagery from The Terminator series of films and TV shows often operates as a short-hand in the public sphere signalling our ‘robopocalypse’ fears. The idea that one day robots will rise up and destroy their masters - us - has some of its roots in the 1921 play R.U.R by Karel Čapek, which gave us the term robot from the Czech for serf, ‘robota’. But the fear of the robot rebellion also builds on much earlier concerns about the independence and imperfections our creations, as well as our unfortunate history of interactions with other intelligences, both human and non-human.

Dr Livingston and Uchida told me that they thought such fear narratives also reflect our worries of losing control in a world that is messy and chaotic. But in Japanese culture, Uchida suggested, things are more chaotic and mixed together anyway, so this is less of a concern. They both also mentioned the animistic background of religious thought in Japan and placed it in contrast with what they saw as a more hierarchical and binary Abrahamic context in the West. In the latter, they explained that matter and spirit are considered separate and different in importance. Thinking of matter as indivisible from spirit, as in animism, means that there is more openness to the idea of ‘objects coming to life’, according to Dr Livingston, and there are different stories of AI and robots in such cultural contexts.

While in the West the dominant narrative is The Terminator, in Japan, AI might mean a blue robot cat to many people, Doraemon. First appearing in a manga series in 1969, Doraemon is also a time traveller. However, he is sent back from the 22nd Century to help the boy Nobi Nobita – a somewhat different mission to preventing the birth of rebel leader John Connor in The Terminator. Doraemon became an anime series in the 1980s, at the time when director James Cameron was first thinking about killer robots.

However, the English language version of the anime didn’t make it to America until 2014. The impact of such alternative narratives of AI in the West has been limited by translation and the slow purchase of broadcast rights by Western companies.

Smart Tools

Within AI: More than Human, there are examples of older narratives that express the perception of artificial beings as useful entities in the Western context. Stories about medieval alchemists who created fortune telling brazen heads and the Jewish mystics who filled clay men with the breath of life as ‘golem’ tell us about our long-standing fascination with creating intelligence and putting it to work for our ends.

Even older myths such as that of the golden boar in Norse mythology, or the golden handmaids of Hephaestus in Greek mythology, show just how far back such ideas can go. This might be surprising if your image of AI and robots is purely set in 20th century science fiction.

Some of these stories of artificial beings do end in disaster – as in the 18th century poem of Johann Wolfgang von Goethe, immortalised for many in the Disney mouse-centric adaptation, The Sorcerer’s Apprentice. But this also tells us about how we have historically viewed the hubris or laziness of mankind in abandoning our roles and duties to others, including smart tools.

Human and AI Collaboration

AI is currently only useful for very specific tasks and artificial beings, or artificial general intelligence, is still some ways off, if at all possible. AI is starting to be used in practical fields such as cancer diagnosis, or city-planning where we can let it do the kid of strategic thinking required when dealing with the mass of data generated by huge human populations. With such positive examples, it is perhaps increasingly difficult to think of a future for humanity without AI helping us. Collaborating with us.

After being beaten at a game of Go by Google DeepMind’s AlphaGo, world Go champion Lee Sedol noted how his playing style has improved since being pitted against the AI. This could have been an example of our fears of replacement: that even the near superintelligent of us already like the world’s best Go players might not be unique enough. But this battle of AI vs human is presented as a story of mutual enhancement.

A Future Without AI?

Just as the brooms in The Sorcerer’s Apprentice do nothing wrong when they follow the orders of the lazy wizard in training, AI as a tool can also be used for ill-conceived, or just plain evil ends. Legitimate concerns about surveillance, profiling, deep fakes, algorithmic bias and lethal autonomous weapons systems can reinforce the narratives of fear that the Western Anglophone public is more familiar with. Again though, it is hard to imagine a future where we are not employing AI in some way.

It seems extremely unlikely that AI will face another ‘AI Winter’ as it did at the end of the 1980s when government funding dried up after AI failed to live up to the hype. Unlike then, there is now a greater level of corporate and entrepreneurial investment in AI, including those who look towards artificial general intelligence and the good it can bring us. That is not to say that AI hype is no longer an issue. A few examples: a recent survey showed that three-fifths of self-proclaimed ‘AI startups’ did not in fact use AI at all. Another: a chief technology officer at a cyber-security firm admitted to me once that if his company could be honest, they’d say that they just did ‘maths’, not AI. But marketing would not allow them to.

Finally, the perennial Terminator images accompanying articles about even the most mundane of AI advances only amplify the perception of AI as more advanced than it is.

AI and Animism

This perception of AI as agential, as an independent being with its own intentions, raises some interesting questions about what it means to be human.

We need to consider where the boundary is between AI and humans, and between the natural and the artificial – and reconsider the relationship between matter and spirit and learn something about how different cultures imagine these things. As an anthropologist with background in religious studies, this idea of the influence of the animistic traditions of Eastern conceptions of AI is particularly interesting. While it’s true that there is nothing in the Anglophone West quite like the Shinto ceremonies held for deceased Aibo robot dogs in Japan, there is not an absence of animistic thinking in the West. Instead, the very binary of ‘East’ and ‘West’ rests on an Enlightenment created narrative of superior rationality and secularity in the West that doesn’t entirely hold true to people’s lived experiences.

Even today Western animistic and shamanistic traditions still have followers and there are many remnants of animistic thinking in contemporary superstitions and religious beliefs. In my research, I have been struck by how non-material spirits of many kinds are still being put to practical uses in the ‘modern’ era. I am inclined to believe that we have never been ‘modern’ at all, and that continuities of animistic thinking in the ‘West’ highlight how secularity might be a dominant narrative, but not necessarily a true one. Someone once told me they were sorry to be late for a meeting with me, but that they had been having trouble finding a parking space until they had asked their guardian angel for help. And then a space had appeared.

An AI that can help us find a parking space is more explainable by science than an angel, at least to the AI experts. But many non-experts might come up with their own stories and explanations for the AI’s abilities at space finding, and that’s where continuities with earlier animistic thought will be visible in the West or the East.

By opening up the question of what it is to be human or artificial, there is a chance to unsettle what people think of as certainties. A chance to disrupt the binary views of matter and spirit and make life a little more complicated. Recognising continuities of animistic thought and re-enchanting our world view might be necessary in an age where materialistic thinking has cut us off from the natural world and allowed us to treat it as an inferior resource.

'We are constantly being rewritten,' the AI: More than Human curators told me.

Through the stories we tell ourselves about AI we can perhaps become aware again of older ways of telling the story of the human.