decolonial_ai_101.mp3: Audio automatically transcribed by Sonix

decolonial_ai_101.mp3: this mp3 audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.

Speaker1:
Welcome to Radical I, a podcast about technology, power society and what it means to be human in the age of information. We are your hosts, Dylan and Jess, two PhD students with different backgrounds researching A.I. and technology ethics. In this episode, we interview Rosiere Buzzy Chetan about colonial de Colonial and post-colonial A.I. and the newly released decolonial AI manifesto. Boosie is an AI policy and ethics researcher and consultant. Her work revolves around ethics, impact and governance of A.I. systems. She combines her lived experience with her interest in postcolonial studies, intersectional feminism and science and technology studies to develop critical thinking about AI technologies and the narratives around it. And today we're just going to get right into it. So without further ado, we're so excited to share this interview with Booth with all of you. Today, we are on the line with Boosie to discuss the I decolonial manifesto. Lucy, thank you so much for joining us today.

Speaker2:
Thank you so much for inviting me.

Speaker1:
So let's just dive right into it. So what is the I decolonial manifesto and how did it come to be?

Speaker2:
What is I decolonial manifesto? It's a it's a big question. I think we will need to. Maybe we will need to start with what is colonial legacy? What is what does it mean? The colonial? What does it mean? The colonial eye? And so on. And. Actually, I think my my answer is going to be a little bit disappointing because I think this is actually supposed to be one of the first topics of conversation that we would like to have as the the team behind the AI manifesto, because we realize that terms post-colonial the colonial decolonisation, the colonial legacy and so on, I mean different things to different people and also the people with lived experiences of colonization, especially. So for some, it is a concrete political project to liberate the land they live in, and for some, it's mostly an epistemological or ontological project. So we actually launched this or put this call out there in order to make space for this plural. And maybe sometimes contradictory visions of what the colonial eye is, or if it is it even possible to imagine the anti-colonial eye or what is left of AI after it's the colonial? So I think this is kind of the purpose of this manifesto to in a certain way. And I cannot. I'm not an academic and I want to also, you know, show respect to all the people who have been researching this for years or also people who live the experiences. I should be part of this conversation. But as somebody who researches who works in policy, advocacy and art in the I or AI ethics field, I can explain a little bit why and how these concepts or these ideas resonated with me.

Speaker2:
So I started learning about AI and I policy in 2017 through an AI policy course. And I remember my first question after the very first class was to go and see the professor and ask about what about the other countries by other, I mean, non-Western because I come from Turkey, right? And this kind of question and thinking about, OK, I understand the United States is the leader and I understand there's China and so on. So it makes sense as the as a continuation of industrialization and, you know, information and communication technologies and so on. But how about the others if this technology is going to have such a big effect, maybe either good or bad, depending on how people project themselves? And then then what about us in a way? And then this pushed me to, you know, look at the intersection of quote-unquote development and and AI. And I put I use quotes when I say development, because that's exactly the point that I that I'm trying to make. Because when I was in, you know, international policy forums about at the intersection of global south again in, you know, like quotes and development and I, I started to feel a kind of unease, although I think everybody was very well-meaning and trying to, you know, do something good. When you look at the, for example, metrics to assess the readiness of a country for AI adoption or development, let's say, and you start looking at the metrics used by international development banks, banks and similar organizations.

Speaker2:
You see stuff like, for example, rule ruler, rule of law index or, you know, like similar things. And when you start to dig deeper at some point, you realize that all these metrics are kind of related to each other. And basically, they also serve for differentiate differentiating Western and non-Western countries. Of course, things are not as simple as that and where there are many statisticians, mathematicians and scientists who are working beyond those. But at some point I remember having seen very like sub indexes or matrixes, I was like, OK, what does that mean? So this is related to this one, they all come together. What I see is like, if you're not following a Western, if you're not a western country and you don't have the western democracy model and you're not at the center of the global political economy as of today and historically, then you are behind, then you're going to lag behind and that's the discourse, you know, anyway. So that was one thing, very deterministic and kind of. Vision, where then Western companies like would come and for for knowledge and technology transfer and, you know, NGOs would try to put out and and invent certain models of management and so on. And the second thing kind that got my attention and that bothered me was that we were trying to in a way promote the potential of AI technologies by saying that the use of AI is going to help Indian farmers. Farmers in doing this.

Speaker2:
But where's the evidence? You know, what do the people in question think about this application? Is it really going to be helpful for them? These were some gaps that I had and and then this pushed me to kind, kinda encountered the work of Art Escobar in development and understand the emergence of development as a as a discipline and the political and economic context that led to it. And I think this was kind of my entry points to think about more critically and to, you know, like have this exposure to post colonialism in a way. Then I was also very much marked by the works of Professor Sid Mustafa Ali. I hope I'm pronouncing his name correctly about whiteness and existential risk, because again, when thinking about singularity and existential risk, there was something that didn't quite, you know, connect with me. And I was wondering what it was, and it was very reassuring to find his words and to have this parallelism between kind of this apocalyptic projections of AI and Singularity and Judeo-Christian ism and the similar themes in this religious and cultural tradition and of course, its implications. Ultimately, what this why this these ideas resonate with me is that. The colonialism or colonial, its ideas about colonial legacy are about putting life and the regeneration of life above reproduction and production and recycling of goods. I think that's the ultimate point why I really resonate with it. But maybe before going to that, we can try to unpack some of the the turns. Or maybe you want to ask that question? I don't know.

Speaker1:
Yeah, that would be wonderful. Maybe we can just do like a quick surface level 101 of what you mean when you say Colonial City de Colonial City and also post colonialism. I think I heard you say that too. And I'm just curious what the difference between especially post-colonial and colonial is.

Speaker2:
I think all in all. And as human beings, we have this capacity of taking the map as the land, you know, and taking, for example, I for whatever we project onto it. So in general, what I understand or what I draw from these frameworks is that we need to look at the foundations upon which this technology technologies or the imaginaries that we also project onto. Those technologies are built and how they are still affecting people who have gone through processes of marginalization and colonization and colonialism, and onto logically how they present, how they feel, how they sense and perceive the world, and as well as our economic and political relations. But just to unpack it a little bit, I'm drawing from again the work of Dr, said Mustafa Ali, where in in a brief introduction to the colonial computing, he unpacks these terms actually very well and very well, and I would encourage the the auditors to go and check it themselves. But basically, when we're talking about colonization, we're talking about the expansionist migration. And example of that is the European expansion migration to what we call the Americas or Australia today. And while colonialism is this this situation of ruling of existing indigenous peoples of so-called new territories.

Speaker2:
So colonialism is the situation of ruling where there is a relation between the centre and the periphery. Colonization is the expansionist movement. But what is not told or what is not? What we don't we don't learn is the. And is there is the actually. The Empire, the history of dark history of colonialism, the empire and the enslavement of peoples and genocide and appropriation of, you know, resources and erasure of languages, practices, ways of being and relating. And this is actually a world post colonialism and the colonial city is, is, you know, drawing our attention towards. And post colonialism. Emerge has emerged in the academia by diasporic researchers like Edward said Homi Bhabha and Gayatri Spivak, who were in Western institutions and who were inspired by the independence movements of formerly colonized states at that time. And this post-colonial studies are also very much ingrained in the English literature department and probably before like. For this reason, they focus a lot on the cultural aspect of by basically saying that, OK, the colonialism, as it is politically, politically, it seems to be over for most of the formerly colonized countries, but the legacies of it are still present today and we are shaped by it and basically also saying that knowledge production is not something neutral, that it's something that is related to power that we should examine.

Speaker2:
And I think Edward cites Orientalism especially has had an important impact on me and, you know, in thinking about it in a way, but how colonial modernity or, you know, the colonial scholars differ from that is that they are also diasporic researchers, but they basically trace the current paradigm of modernity to the early European conquest of what we call today Latin America. Basically, Christopher Columbus, so-called discovery in the 15th century and so on. They basically they say that this is the modernity paradigm that we live in today is the continuation or the culmination of a 500 years history of enslavement, empire, colonialism and so on. And. Yeah, I think all of these frameworks, frameworks and thinking kind of urges us to not be myopic in a way and to look at the history and the foundations and of the ideas that drive IE where they coming from and how they have impacted people and still impacting especially marginalized communities or anybody. Actually, that is beyond the norm of the universal default western wild white male in a way.

Speaker1:
For folks who haven't seen the manifesto yet or or read it, could you just say what it is? Quickly, just like on a very basic level, or at least like how you how you consider it?

Speaker2:
I think the manifesto is an invitation, and it's I think it's also invites us to to have some space for irony. And sort ability to hold contradiction. And how I'm going to explain it because a manifesto usually is a powerful political tool, right, where one one person or like a group of people say this is how things should be and they seek political impact and change in a very well defined terms. So the manifesto to for for me to give a little bit background as far as I know, has come from this observation around the consolidation or consensus around AI principles. So the you know, we had this kind of not the emergence, but this this hype around AI, you know, starting in 2016 and and that followed with the proliferation of AI ethics principles by companies, international organizations, you know, and many, many different actors, really. And this was for some is called as the first wave or like first kind of phase of AI ethics as of today. And. The where did principles come from? You know, they mostly western right and the ethics that we're talking about in the principles, they're mostly Western ethics, utilitarianism counts and, you know, like other frameworks of Western ethics and people. Also, some people also argue that now like, OK, it is time to move to action and operationalize ethics principles, and therefore these are the processes. These are the audits and so on and so forth. So we see a kind of consolidation consensus.

Speaker2:
I'm not sure about consolidation around these principles, especially around certain intergovernmental organizations such as the OECD or new ones like Global Partnership on AI. And again, with the center in the in the Western Center and kind of, you know, at other actors or non-Western countries following suit or not. And so I think the need for this conversation emerged from from that that, OK, things are moving fast and there is a consolidation. And it sounds or seems very familiar. It seems like something that will then, you know, it seems like. Things are going to build on top of it, so we need to make sure that we stop and we also create space to discuss and still question how about we thought, how about we thought about this differently? How would we think about AI ethics if we took a non-Western ethical framework, such as, for example, Ubuntu and Sabelo, who is part of the team and one of the co-chairs, has done amazing work on that. You know, because in Western ethics, you define the individual as one person and you equated with rationality. But in the Ubuntu framework that he presents, a person exists through other people, other persons. So we are it's the relation and not one person or one person's, you know, atomized relationship to any other thing that's at the center. So I think it was kind of coming from this urge to, you know, say, just stop and we need to still keep having this conversation.

Speaker2:
And how can we maybe try to facilitate or enable in a way, a vision of the universe where there is, you know, different localized centers of discussion and knowledge? So in that sense, the manifesto, it is not called a manifesto in the traditional sense, but manifesto it suggests it wants to use the strength and the dynamism of a of a traditional manifesto in order to, you know, like invite and and and call for action and discussion and dialogue. But it also deeply recognises the need to hold plural views. A contradiction and not necessarily consensus, because we don't as human beings again. I feel like we don't have a lot of patience for that. Our kind of maybe colonial legacy, like colonial mind wants to be sure of things, wants to have ready-made answers. Yes or not, this or that, OK, how do we move at the same time? This doesn't mean that we have time because at the you know, the impact like AI technologies are already widespread and especially in very critical areas of our lives, right? And not everybody impact is impacted by them, equally so. So this is not to discourage any kind of action, but we think that we can have this keep having these conversations in parallel and change also how we measure things. And I think this echoes very well with the measurement in the series that you have had at the podcast recently.

Speaker1:
So something that we try to do on this show whenever we have an opportunity to is to take these conversations and more abstract ideas and to make them a little bit more concrete by applying them to a case study or a real world example. And so something that stuck in my mind when you were explaining this manifesto is right now we have all these AI systems that are attempting to apply these AI ethics principles in a really Western centric way, like utilitarian ethics, for example. And now we have the potential to apply other ethical frameworks that are less Western centric, such as Ubuntu Ethics. And in my head, I feel like I am experiencing this cognitive dissonance because I recognize that with large tech, there is usually an issue of scale. And so sometimes, even though people have different preferences and ethical needs in different locations, geographically, culturally, whatever it is, if we're scaling these systems, we can't always pick and choose based off of what people need. Sometimes there needs to be like one solution for everything, at least the way that AI is created. So I'm just curious, like with that case study in mind and the idea of scale. How does the decolonial manifesto maybe help us interrogate some of those challenges that we might experience so that we can come up with? Or so that we can utilize some of these non-Western centric ethical principles in our AI systems?

Speaker2:
Great question. I think there are multiple things. First of all, I think the manifesto or the AI is not an attempt to make AI more inclusive or make AI better, or to offer a tool for big tech companies to, you know. I mean, ideally, of course, you want everybody to be to be asking these questions. But yeah, so what I'm trying to say is that it's not an attempt to make better I and if that's the case and if that's possible with these ideas and frameworks in mind, that's great. But I think what I'm trying to say is that the the idea of refusing. Certain AI technologies and applications should always be there and possible. You know what? This is what I was saying at the beginning as so what is left of AI? When, when, once is, let's say, the colonial or the colonial, the colonized? Maybe there isn't a lot of things left because it is already built on an infrastructure of of colonial and power. So I think at the end of the day, it's about understanding and examining also power and thinking about the question of scale independent of this oligopolistic structure of economic concentration, you know, unequal extractive relationships in the global supply chains or where the materials, data and labor for AI to perform comes from. It becomes complicated, we cannot think of scale without having this in mind, and I don't have a have a solution for that, but I think the question of scale is part of the problem and something that we need to reconsider and think about. Ok, how do we make sure that what we develop actually serves people or how people can themselves benefit from the benefit from these technologies can, can offer or not? Or, you know, maybe that's not the solution. The solution is something that we already have access to, and it's not about the technological solution and so on.

Speaker1:
So one of the things that I was really struck by when when reading through at least the introduction of the manifesto is looking at the challenge posed by language that we use to talk about AI. And you've spoken a little bit about this. But in this context, where so much of technology has been dominated and designed by Western male voices, whiteness and wealth, and I'm quoting from that introduction. You all seem to want to challenge those voices and also the technology that we have now been socialized to to use, and so I'm wondering how we can disrupt. Maybe that's not the right word, so correct me, but how we can upend some of that language when it's the. To some degree, the only language that we know how to use as a society when we talk about AI

Speaker2:
Language is a tricky topic. Yeah, I think first of all, what comes to my mind is this question of English, for example, you know, speaking in English, like the fact that I now speak about this in English, and I wouldn't be able to hold this conversation in my mother tongue, which is Turkish. That's a contradiction that I have to hold, and that makes me feel sometimes alienated. And the second thing is, I think language has a has a double use in the sense that when. Again, drawing from my personal experience when I was like working in policy and like at the intersection of development and policy and I and feeling this sense of unease that something was not quite right, not having the language for it, you know, was a problem. And like encountering the works of Quijano or Dr Paula, requited Chicano and other scholars that I look up to in this field was empowering. But at the same time, we need to think inside and outside language and the categories and mental models that we build in the language and also how we construct them in a way that are these categories and concepts that we we construct are the building hierarchies and power asymmetries, as it is kind of the case for what these frameworks are, are criticizing in a way.

Speaker2:
Or can we look at them differently so language can be? Um, can be problematic, for example, we can think about it as the, you know, AI ethics language that is not maybe appealing to some people because it doesn't encompass the questions of justice again, you know, like historical processes of marginalization and inequity and so on. So. I think we need to be able to be very contextual and choose our words, according to the context. We are in thinking about also what resonates with people and also not be afraid of moving along certain words and concepts and reinventing new. Because for me, at the end of the day, ideas or like words are things that we assign meaning to. But the essence is something that is not maybe easily captured by words. It's also, you know, decolonial. It is also, for example, a sense of feeling and being. And it's a praxis rather than, you know, these things that we put on a map and, you know, accept as like a source of truth or it's fluid in a way, this is how I see things. So I don't know if that's that answers your question, but I think it requires us to kind of establish a certain kind of distance and irony with the words that we also use and be open to maybe create new ones, depending on the purpose.

Speaker2:
If we want to be understood widely, how we want to be understood and so on, and like thinking inside and outside language. And also, I think it's important to create spaces to have this kind of conversations in other languages than than English. And we have a Spanish version of the manifest on the website, and we're also aiming to have other other versions and I say version for lack of a better word. It's also English, also visibly not my first language, but yeah, maybe, probably. Maybe, you know, people are not going the going to engage with this topic in another country with these terms of the colonial. I don't even know how to translate this into Turkish, for example, but maybe through stories and narratives and maybe by remembering our like practices, spiritual practices, you know, like local stories and so on. And that's actually the aim of this manifesto in a way not to promote it like readymade certain concept, but just to, you know, launch a call to make different, simultaneous, different but similar conversations happen. And maybe like create a universe of decentralized, yeah, culture locations.

Speaker1:
Absolutely, yeah, and thank you so much for the work that you and the team have done on this manifesto to start these kinds of conversations like the one that we're having right now, and for other people who are hoping to also have similar conversations or listeners who are resonating with everything that you're speaking about here. Where is the best place for them to go to engage with this manifesto? And also, I think there's a way for them to sign it to endorse it. Is that correct?

Speaker2:
Yes. And so where they can find the manifesto is. Manifesto, but I. And it is called the decolonial manifesto, and they can also sign it, and they can be part of the mailing list. And what we're intending to do with it is that we'd like people to just feel empowered to, you know, just exchange and start a dialogue and also organize workshops. Or any format that they imagine around also, I think making other visions possible or, you know, palpable through arts stories. Yeah. Other other ways of relating.

Speaker1:
Absolutely. And for listeners who are listening with their ears and not reading with their eyes, if you didn't catch on yet, this is manifesto manifesto spelt with a y manifesto made and why f s t o II is that website. And that's also in our show notes, along with a bunch of other relevant amazing resources that were mentioned and not mentioned in this conversation that are relevant to the conversation. But for now, we are out of time, so we say we just wanted to thank you so much for coming on our show and having this conversation with us and for the really important work that you were doing out there.

Speaker2:
Thank you so much for having me. That was a real pleasure to have this conversation with you. Thanks for your amazing work to.

Speaker1:
We want to thank you again for joining us today and for this wonderful conversation. I think the thing that I'm most stuck with or struck by is this concept of language, and especially when Bruce was talking about how in Turkish you couldn't even be having this conversation and the level of translation that needs to happen in order to actually have these conversations about AI in general, across time and across space. I haven't really thought a lot about the language that we use and the different meanings and symbols that are inherent in that language and how they pervade whole systems of power and create and recreate systems of power. And so even the fact that this podcast is in English, it creates some barriers around how the information and the language of the guests that we have on the show, like where it can go and where it can't go. And so thinking about that, even as we have technology where one of the narratives is like, you know, the internet has crossed boundaries in some ways, sure, but in other ways, it has also created boundaries between cultures, between peoples and between different sectors of language in the world. Absolutely. That was something that really struck me in this conversation to, I mean, the idea of not being able to speak about your job and your research and maybe some of the things that you're passionate about in your mother tongue.

Speaker1:
That's something that I take for granted as somebody whose first language is English, primary language is English. And it's something I haven't actually thought about or really reflected on as deeply since a few years ago, when I was doing a little bit of work teaching computer science in Colombia, the country, and I was teaching it in Spanish. I had this realization for the first time that so many of the languages that we code technologies in, like computing technologies, are built on a foundation of English. And I mean, there's like, for example, the idea of like four loops and if and else conditional statements like those are English words that have been put into a computing language as keywords that are in the English language. And so something like computer science that for so long the way that I was taught it in school in the U.S. was something that is like this global skill set than anybody can have and anybody can do. But there's actually this huge barrier to entry in this gatekeeping that exists because for people who don't speak English at all, the idea of learning how to code and how to learn computer science even in languages that aren't in English might still require quite a bit of English knowledge. And so that was something that was really eye-opening for me at the time, and I was reminded of that in this conversation.

Speaker1:
One other thing that stood out to me was the difference that they talked about between this model of inclusivity versus what they're doing. And pretty clearly in the manifesto, it says, we do not seek consensus. We value human difference. And I feel like even over the course of this show, we've grappled with what inclusivity means and whether it's a good thing. I guess whether it's a value neutral thing, inclusivity for whom, when, et cetera. But the fact that, you know, returning to this language point, I guess the fact that the manifesto says, you know, we reject the Western normative language of quote unquote ethical AI and suggestions of quote unquote inclusivity that do not destabilize current patterns of domination and address power asymmetries. And it does make me think about different instances. And, you know, I don't want to to call out industry directly. I think there's a lot of different systems where this can be applied. But, you know, inclusivity statements or, you know, we talk about different dog whistling or just like those checkboxes of inclusivity when they're not actually substantiated and when they're not actually doing the work of upending the systems of oppression that exist, but that they just exist in order to include the word of inclusivity because the word of the day.

Speaker1:
So the fact that they're able to bring out so much nuance beyond consensus and beyond concepts of inclusivity, I think is a very powerful and an important statement to bring into this conversation. You're also reminding me something that I definitely wanted to mention in this episode is that this interview and this topic is long overdue for this podcast. It's something that we've been wanting to talk about on the show for a really long time, and we were just waiting for an opportunity. And when this decolonial II manifesto came out, we jumped on it because we thought this was a really good case study to ground this topic in. But as for, I mean, many of you who are listening, if you've been around since near the beginning of this podcast, you might remember that we used to ask guests on this show for the first six or seven months or so of these interviews. We would ask them to define what radical I was to them because we were trying to work with the community to uncover what the different potential definitions for radical could be as it is applied to AI and for this project, that was a part of our initial goals with this project. And one of the topics and themes that we continually came back to with some really incredible scholars in this field was this idea of like radical as the root and digging into the root of these problems in society and trying to get past the surface level to really understand what are the causes of inequality and oppression and discrimination.

Speaker1:
And some of these really sticky social challenges that social problems that we have not as a society found answers to and that we may never find and likely or never to find answers to. And so this this conversation about colonialism and colonial and post-colonial just fits really, really well into that narrative that this this entire podcast was built off of in a way. So for me, I'm I'm being reminded of some of the roots of our own mission with this project, and we do invite all you listeners out there to read the manifesto. And if you feel so cold to sign and and to continue this very, very important conversation. But for now, for more information on today's show, please visit the episode page at Radical A dot org. If you enjoyed this episode, we invite you to subscribe, rate and review the show on iTunes or your favorite pod catcher. Catch our regularly scheduled episodes the first Wednesday of every month with some bonus episodes in between. Join our conversation on Twitter at radical iPod and as always, stay radical. Typekit.

Sonix is the world’s most advanced automated transcription, translation, and subtitling platform. Fast, accurate, and affordable.

Automatically convert your mp3 files to text (txt file), Microsoft Word (docx file), and SubRip Subtitle (srt file) in minutes.

Sonix has many features that you'd love including automated transcription, automated subtitles, world-class support, share transcripts, and easily transcribe your Zoom meetings. Try Sonix for free today.