皇家华人

Professor Harris describes his research as having become increasingly relevant in recent months in light of the UK and US elections. From anthropomorphism and dehumanisation, to AI and deep fakes, the subjects in which he specialises are hot topics of public debate.

His real interest is in understanding how our minds create our reality. This led him to forge an academic career as an expert in psychology and behavioural science.

鈥淚 think what's interesting about behavioural science and psychology is that we get to study things that no one can directly see,鈥� he says. Behaviour, he explains, is really just an 鈥渙utput鈥� of the mind.

Much of Professor Harris鈥檚 work involves studying how we perceive other people and our ability to imagine what others are thinking. 鈥淚t turns out that we don't just reserve that skill for other people though. We also do it to things that aren鈥檛 human,鈥� he says. This is where his interest in anthropomorphism comes in.

Professor Harris looks at how we perceive non-human objects, like AI, as human beings.

He is particularly intrigued by the moral dimension of anthropomorphism and AI. An example he gives is a self-driving car. If it kills a person, who do we hold accountable for that behaviour? Is it the manufacturer? The person who owns the car? The car itself? 鈥淵ou get these really interesting ethical questions that arise around agency,鈥� he says. 鈥淓ven though we can view AI in human terms, if the humanoid robot you were watching gets run over, you're probably not going to shed a tear.鈥�

One of the key reasons we instinctively separate humans from AI, Professor Harris explains, is that when we engage with other people our brain waves and physiology spontaneously synchronise. Even our pupils dilate to match those of other people.

This doesn鈥檛 happen with AI, at least not yet. Things are changing though. Professor Harris uses brain imaging to look at the most recently evolved parts of the human brain, known as the neocortex. With the use of technology, he says, we could lose some of the processing in these regions. 鈥淥ur hyper-social nature means that our brains are built to interact with other people. But the way we interact now doesn鈥檛 fully tap into this mechanism. For those of us who interact with others via devices frequently, it's a huge issue. There may come a point when we won't be able to tell humans and AI apart,鈥� he warns.

Professor Harris is also worried about the impact of the erosion of the boundaries between real and digital life on society.聽鈥淚 think we live in a very interesting moment because it seems like we're in this movement where reality gets called into question constantly as part of political debate.鈥� In this post-truth world, people鈥檚 reputations are shaped by the groups they're aligned with, not necessarily what they've done.

There is hope though. While the media may be increasingly encamped with political parties, Professor Harris doesn鈥檛 believe that AI-generated deep fakes are likely to sway elections. 鈥淚 think people are much more savvy users of the internet than we give them credit for. Every time we get information through the internet, we know there's a lot of junk out there, so we鈥檙e not just naively consuming stuff 鈥� we evaluate. Is this real? Is it fake?鈥�

Professor Harris鈥檚 main concerns are the lack of regulation from government and the danger of AI鈥檚 inbuilt biases. He points out that currently AI companies have free reign to do whatever they like, and we lack control over how our data is used. If AI is essentially a big correlation machine, then what we feed it determines what it produces. 鈥淎I has an idea of what something is, not because it experienced it in the world, but because we told it that's what the thing is,鈥� he says. But not everyone is part of this new AI future. As Professor Harris explains, if you live in a remote village and don鈥檛 create online content, then your experiences are not influencing AI鈥檚 algorithm. Discrimination then becomes more entrenched.

It may sound like science fiction, but Professor Harris believes that in the future AI will gain similar perceptual mechanisms to聽those of humans. His question then is what AI culture might look like: 鈥淎s humans, we use culture to help interpret the world. Will AI culture be a replica of ours? Or something entirely new?鈥�

However, his main concern is with the present: 鈥淲e鈥檙e living in a period where reality is becoming subjective,鈥� he points out. 鈥淯ltimately we should focus on the issues facing us right now聽such as the climate crisis, political polarisation, and entrenched inequalities聽rather than worrying too much about technologies of the future.鈥�

YouTube Widget Placeholder

Related

• UCL Division of Psychology and Language Sciences