The Good Robot

Kanta Dihal on Decolonising AI Narratives

August 10, 2021 University of Cambridge Centre for Gender Studies
The Good Robot
Kanta Dihal on Decolonising AI Narratives
Show Notes Transcript

In this episode, we chat to Kanta Dihal, Senior Research Fellow at the Leverhulme Centre for the Future of Intelligence who leads the Decolonising AI project. We discuss the stories that are being told about AI, why these stories need to be decolonised, what that means, and how we should go about it. We discuss the need to examine a plurality of local stories about AI and Kanta recommends us her favourite science fiction narratives from around the globe that are challenging the supremacy of science fiction from the Anglophone West. 

KERRY MACKERETH
Hi! We're Eleanor and Kerry. We're the hosts of The Good Robot podcast, and join us as we ask the experts: what is good technology? Is it even possible? And what does feminism have to bring to this conversation? If you wanna learn more about today's topic, head over to our website, where we've got a full transcript of the episode and a specially curated reading list with work by, or picked by, our experts. But until then, sit back, relax, and enjoy the episode.

ELEANOR DRAGE
Today, we’re talking to Kanta Dihal, who's a Senior Research Fellow at the Leverhulme Centre for the Future of Intelligence and who leads the Decolonising AI project. We discuss the stories that are being told about AI, why these stories need to be decolonised, what that means, and how we go should about it. We discuss the need to examine a plurality of local stories about AI and Kanta recommends us her favourite science fiction narratives from around the globe that are challenging the supremacy of science fiction from the Anglophone West. We hope you enjoy the show. 

KERRY MACKERETH 
So thank you so much for being here with us. Could you tell us a little bit about what you do and what brings you to the topic of feminism, gender and technology? 

KANTA DIHAL
Sure, thank you so much for having me. So I'm Kanta Dihal. I'm currently a senior research fellow at the University of Cambridge, at the Leverhulme Centre for the Future of Intelligence. And I lead the project, ah, ‘Global AI narratives’ or ‘Decolonizing AI’. And more generally, my research focuses on science narratives. So how narratives shape science, how science shapes narratives, and how together they shape the perceptions, expectations and resistance to science and technology. And my research has a really intersectional focus. So I look particularly at the point at which the relationship between the portrayals and the technology breaks down. And that's often at the point where marginalised groups come into contact with narratives and technology that are not their own. And that, of course, includes issues of gender. So people who do not see their identities, their bodies, their genders represented in either the narratives or the technology, as they're, of course, at risk of further marginalisation and harm.

ELEANOR DRAGE 
We’re The Good Robot, so we'd like to ask what good technology means to you? Can they even be such a thing as good technology?

KANTA DIHAL 
Yeah, that's a million dollar question. It's, I think it's a question that really shows what the problem is with our technology today. I do think we can have good technology. And in fact, in my work, I take a really broad view of what technology is, because I find that there’s a lot of value in the concept of technogenesis, which is the idea that humans have evolved together with and through their technology. So technology has made us what we are today, and it's easy to see how that involves using technology for good and for ill. I think the clearest examples of good technology can be found in medicine, particularly the extent to which infant mortality can be avoided these days. But I think modern medicine also shows that often, the problem is in getting to that point where the technology is good. The development of the technology in its earliest stages is often where most harm is done. So in the context of medicine, it is experimentation, or only including in the treatment certain groups, certain people who can afford it, testing only on, say, men, so that that same medical technology might in fact, be harmful to other people, other bodies, other genders. And as things develop further, it is often resistance to those flaws and resistance to those harms that really lead to improvements. And I think what we really should be aiming for is instead that the development of technology itself becomes good, as well as the technology that eventually comes out of it.

KERRY MACKERETH   
That's fascinating. And I'm so compelled by your insistence on the process and where these technologies come from, and not just where they’re going, which can be such an emphasis in this very futuristic kind of language that is used around technology, especially AI, and I have to admit, I'm hugely interested in the project that you're the lead on, Decolonising AI, I think it's so important. And I was wondering, I'd like to hear why you think stories are so important to this broader project of Decolonising AI?

KANTA DIHAL 
Yeah, well, as you said, it's often a narrative looks forward with regard to new technologies. And in the context of artificial intelligence, these forward looking narratives have really shaped perceptions and expectations of the technology, what it is, what it looks like, what it should look like and do, who it's for. And that means that all the biases and preconceptions of stories sometimes written more than 100 years ago, and stories nearly always created by a very narrow demographic, usually of a North American or British male white science fiction writers. And because those stories have been around for so long in such a narrow form, and from such a narrow demographic, they have perpetuated these biases that have led to the marginalisation of many groups and parts of the world. And this is why these narratives matter in the context of decolonisation, when technology from the same parts of the world from North America, from the UK, are again being sent out to the rest of the world, often with very explicit white saviour narratives of ‘look at how good my technology is for me so it should be great for you!’. Those narratives are accompanying the technology, ‘look at what we have imagined this technology could become’. Whereas very often, their response to that is ‘Oh, actually, that narrative is really kind of imperialist, and sexist. And actually, that's not the future that I want. So why would I want this technology?’ And so therefore, in order to decolonise the technology by which I mean that those imperialist systems and impulses and ways of thinking about the technology, in order to take those out, we must also take, take apart the narratives that accompany them.

ELEANOR DRAGE   
You lead the Global AI Narratives research project, which have hosted some incredible workshops with research groups from around the world. So can you share with us some insights into global perspectives on AI?

KANTA DIHAL
Yes, I think the most interesting perspectives that I've seen are, I guess I wouldn't call them global, but local, in the way that they really are grounded in different parts of the world, in different systems of thinking, and show precisely how the narratives that in the West people grow up with are also local in their ways. So thinking of things like the robot apocalypse, and how, how common a narrative that is, especially in the Anglophone West, looking at the kinds of narratives that are prevalent in the rest of the world really shows how limited of view of life with intelligent machines that is. So for instance, we have seen that in Japan and China in particular, ancient philosophies about living with other intelligences, different ways of understanding what another being another soul, another intelligence is, really prepares people differently for life with intelligent machines. Because from those perspectives, even if we were to create an artificial general intelligence, a machine that is as intelligent as a human, that doesn't mean that suddenly, there are two intelligent species on the planet, humans and machines, who then contend for supremacy. That is very much an Anglophone Western narrative. Instead, if you look at it from a philosophy that has always imagined humans as one of many different ways of being, of being intelligent, of being connected to each other, of being an embodied, soul, intelligence, power, whatever you will call it, then it's much easier to fit an artificial general intelligence into that framework. And then your understanding of that does not have to be that it will contend with us for supremacy. Because we do not contend for supremacy with the water, the mountains, trees, although it again, in some perspectives, some people feel that we should or we are.

ELEANOR DRAGE
This is a really important point about giving space for heterogeneity of local knowledge in visions of the future expressed by AI. And to me, it expresses the idea of situated knowledges, which is a key foundational idea of feminism. So I wanted to ask you, what does feminism mean to you? And what does decolonisation mean to you, and how do they relate?

KANTA DIHAL
Yes, I mentioned earlier that my work is fundamentally intersectional. And I just want to emphasise that again, because I think this is a term that should be understood more widely in the world of AI, especially now that the field of AI ethics is really burgeoning. And there is a lot of attention to things like AI and gender. But these things are, are still quite often understood in a non intersectional way. Sometimes even in a very narrow sort of ‘women in AI’ or ‘women in tech’ way. Whereas, of course, intersectionality, the coming together of different forms of oppression and understanding that those can exacerbate each other, that one is not simply a woman, and a person of colour, but one is a woman of colour, which is a different way of being that brings together being at the receiving end of many different forms of oppression, some of which are specific to being a woman of colour, rather than being these separate things. And the same goes for things like gender, and disability, and gender, and class, race, and class, disability, and class. All of these form these systems and these networks of oppression together. And this is why my decolonising project is intersectional, is feminist, but also all the narratives work that I do outside of the project of decolonizing AI.

KERRY MACKERETH
So fascinating! Thank you so much. I'd love to hear a bit more about how then your intersectional and your feminist approach manifests in the work that you do on narratives, both in the sort of Global AI Narratives projects, but also outside of that.

KANTA DIHAL
Yes, indeed, so looking at these different factors, when taking into account how stories about artificial intelligence are being told, and in different parts of the world. And how then AI is perceived in, in those areas, in those cultures, really brings out these points of potential improvement, learning where we could learn from each other, where we could understand each other better, where we can understand the consequences of this technology better. Thinking about parts of the world that do not have this perception of artificial intelligence, as ‘it's going to take over the world, it's going to kill us, it's going to be bad for us’. Those also very often then see AI as, as a potential helper buddy, so a tool that can improve our lives. But then looking at those narratives from a gendered or a racialized perspective can bring out some very interesting aspects that might not at first be visible, because the narrative of say, a care robot whether it looks humanoid, or like a seal or not, in Japan and South Korea, these, these narratives are very widespread and very positively received. But underlying that is, first of all, a gender issue, the, the role of care work, the role of women's work and how it is being appreciated, how it is being evaluated as replaceable by a machine. But also, rather insidiously, there is an ethnicity dimension where the major rationale for introducing care bots is the idea that there aren't enough people to take care of an increasingly ageing population. But we of course know that there are plenty of people on this planet, it is just that these people for one reason or another are not considered good enough, not considered suitable enough to take care of the ageing population of another country. So there is a very strong nationalist and racist narrative underlying these apparently positive narratives about automatisation, about robotisation. And so looking at narratives about AI through these different lenses brings out these underlying paradigms that aren't visible at first sight.

ELEANOR DRAGE
Can you give us some examples of your favourite stories about AI that you think are doing feminist decolonial work?

KANTA DIHAL
Yes, um, there are, there is some really exciting work going on. This is I think, one of the most important things I guess that I want to get across with the AI narratives and the Global AI Narratives work, there are often calls for new narratives. But in fact, what we should be doing as well is give space to the narratives that are there but that are not listened to as much as you know, a whole range of 100 year old narratives, all coming from the US. And I really like the ways in which, especially English language, science fiction, is catching up with that. Because, you know, English language science fiction, for the longest time, was a rather conservative male bastion that was being broken up in the 1970s and 80s by the new wave, which included amazing writers like Ursula K. LeGuin, and Octavia Butler. But after that, it did seem to turn back in on itself for a while, and especially with the mainstreaming of science fiction in Hollywood, which led to a whole range of really rather conservative views of science fiction, really conservative AI narratives. In the past decade or so, publishers such as tor.com, and some of the biggest science fiction magazines like Uncanny really have thrown the gates wide open and there is such exciting work coming out of that. I particularly like Annalee Newitz’s novel Autonomous, which is an amazing novel about a robot, an intelligent autonomous weapon that struggles with its gender identity, and first develops a male gender identity and then develops a more female gender identity. And so, written by a transgender author, I think this is a perspective that is just so much richer and more interesting than the umpteenth story of man falls in love with his sex bot. But also stories that look at the consequences of modern technology in other parts of the world. So we had a workshop on AI Narratives in India, where we had a range of science fiction authors, Samit Basu, S.B. Divya, who have written great novels about the implementation of different technologies that we have nowadays and their future extrapolations, but set in a very different part of the world that deals with technology very differently. So S.B. Divya just had a book out last month, Machinehood. And I'm also thinking about P. Djèlí Clark, Trinidadian American author. So P. Djèlí Clark is the author of the short story, A Dead Djinn in Cairo, and now set in the same universe this year, he has a forthcoming novel, A Master of Djinn. And I particularly enjoyed how he has a parallel Cairo in the 1930s, with a very awesome female protagonist, who, who deals with colour automata, with clockwork beings that are very much grounded in more Middle Eastern perception of the history of automata.

ELEANOR DRAGE
Fantastic, well, you've given us so much to read, I'm really excited to look through some of these, and we're going to attach a reading list with this episode. So we'd love it if you could add those to it, we can all check them out. I did my PhD in science fiction written by women so I think it's fantastic that we are looking back to some of these narratives that were unable to be read because women were writing for example, using their husband's pen names. There's a Italian writer of science fiction called Roberta Rambelli who wrote as Rocky Docson, Hunk Hanover, Joe C. Karpatí, and Igor Latychev, and of course, that excavation work was so important to uncovering these wonderful magical narratives that gave a slightly different idea of the world, what it means to be human, what it means to be an alien. And part of this amazing, extraordinary work that you do is also uncovering these narratives and allowing current writers of science fiction, women today, to trace other genealogies of writing science fiction that aren't readily available. So can you tell us then, what's the future of AI narratives? What have you got in store for us, what other projects are coming up? What should we be paying attention to?

KANTA DIHAL 
Well, the global AI narratives project is continuing this year with workshops in Sub Saharan Africa and in the Czech Republic, after which, we will be launching a report on how our work in the Middle East and North Africa. So the work on AI narratives in that region, which we have done in collaboration with the American University in Cairo, will be launching at the UN AI For Good Global Summit, which is also where we started our work. So it comes nicely full circle. After which we are working on a book on Global AI Narratives, which hopefully will form a kind of a sequel to the book AI Narratives, which came out with Oxford University Press last year. So far, it looks to be incredibly exciting, and it involves contributors from all over the world. That while the book itself will be published in English, that will require some translation for several of the contributors who will be writing their contributions in their preferred language, which just really shows the reach of the project that I'm most excited about, that it manages to reach beyond even English speaking academia, which is often seems to be just sort of assumed that that's academia. Whereas in fact, of course, there are, you know, traditions in academia in other languages, countries, parts of the world, where amazing work is going on that just completely slipped under the radar here. So that's what I'm most excited about is bringing out those perspectives.

KERRY MACKERETH   
Eleanor and I can't wait to see all the exciting things that come out of these workshops that you've been doing, and of course, the publications that will soon be released. So thank you so much for joining us. Once again. It's been really, really wonderful to hear about your work, and we can't wait to hear more about that soon. 

KANTA DIHAL
Well, thank you so much for having me. I'm really excited to be here and with such amazing hosts and other guests you've had on this podcast.