I do feel that AI is in many ways actually a response to a general context of collapse. I think the optimizations of AI as it manifests socially and economically are actually pretty well suited to a sort of neoliberal status quo, because that is about very much about a sort of reductive isolating essentialism, you know, in which things can be optimized.
SPEAKER_00I'm your host, Paris Marks, and this week's guest is Dan McQuillan. Dan is a lecturer at Goldsmiths University of London and the author of Resisting AI. There is growing discussion about the harms and the drama of social media, of the amount of time that we spend on screens, and certainly of the generative AI tools that are proliferating through our lives, especially in the past few years, and how we are going to respond to all these things. We know that there are costs that have come of the way that these digital technologies have transformed our lives, in particular, the way that these companies have sought to implement them in workplaces to suppress the rights of workers, certainly in government to reduce access to public services or make them more discriminatory through the argument of efficiency. And that's not to mention the environmental costs that have come of all of this, especially as we're seeing these generative AI tools proliferate despite the amount of computation that they require. And so the question is not just how are these technologies transforming our society and how would we potentially use less of them, but what is a different way of understanding these technologies and of thinking of a politics that would change our society for the better while reducing our dependence on these computationally intensive tools that seek to capture our attention and make us dependent on them? And so that's part of the reason why I wanted to have Dan on the show today, because he has been writing a lot recently about this concept of decomputing. And I wanted to understand that better. What does it actually mean? What is the politics behind it? What is it trying to achieve? And what does a better world actually look like where we start to critically interrogate the type of computation that we use, the type of digital services that we use, and are much more focused on technology for public good rather than technology that serves shareholders and the ideological visions of the people who lead the tech industry, particularly in the United States, but also in other parts of the world as well. Our governments are really letting us down on this. And so if we want to try to move a different vision forward and to pressure them to do something different, we also need to have a vision in mind. And that is what I think something like decomputing might help us to consider. So this is probably a bit more of a thought-provoking episode, thinking about a bigger idea than maybe something that we usually talk about. But of course, we ground that in things that are definitely happening in issues that you have been seeing us talk about on the show in the past. So I think that you'll enjoy it. And if you do, make sure to leave a five-star review on your podcast platform of choice. You can share the show on social media or with any friends or colleagues who you think would learn from it. And if you do want to support the work that goes into making the show every single week, so I can keep having these critical, in-depth conversations that explore the ways that technology is impacting us and our lives and the world around us. You can join supporters like Peter from Dawson City, Yukon, Carol in Portland, Oregon, Luke in Leiden in the Netherlands, and Lindsay in Seattle by going to patreon.com slash TechWon't Save Us, where you can become a supporter as well. Thanks so much and enjoy this week's conversation. Dan, welcome back to Tech Won't Save Us. Fantastic to be here. It's really great to talk to you again. Obviously, you've been on the show in the past and you were in our series on data centers, I believe it was, last year to kind of dig into all these things. But I wanted to have you back on the show because obviously we've talked about different aspects of AI and data centers and these technologies, of course, based on the book that you wrote, Resisting AI and some other work that you've been doing since then. But you've also been working on this concept of decomputing that I find really interesting. And so I wanted to dig into this with you, but obviously there's kind of a broader context that we should talk about before we dig into that specific concept, right? To know what we're actually responding to. And I feel like a good place for us to start is actually what's happening in your country right now, because it seems to be a good example of how governments are just absolutely chasing this AI hype without any consideration of the potential consequences of this. So, what do you make of the labor government's pursuit of AI and kind of the policy approach that they have developed to try to attract AI companies and tech investment?
SPEAKER_01It is interesting, and I am using air quotes there, because obviously, in a lot of really important ways, this sort of quite florid manifestation of technologies overlap with fascism is most visible in the USA right now, most probably. But the UK is a really interesting case study because you know there's a political party in control without any real opposition at the moment, and that's what will happen in a few years, is another story that's tangled up with this. But at the moment it can it can basically pursue a line. And unlike the last government, which were basically just a very familiar bunch of tofts and crooks, this lot are much more ideologically unified. Um the problem there being, that this is my kind of understanding, my or rather my assessment, that they don't really have an ideology per se. You know, they called the Labour Party, and many years ago they would have had uh people within it would still, some people within it would still see themselves as being, say, committed to socialism and committed to the welfare of the ordinary person, so on and so forth. I mean a minority, but a number of people would still do that. But as a party, I mean it's pretty clear they abandoned, have completely abandoned it. No, I'm I'm laying that groundwork because they are completely AI pilled. And that's what leads me to say those things. I mean, you could look at it the other around and say, this is a government that is, you know, so so kind of cleanly and purely AI pilled. They they they it's almost like they exist only to sort of form a uh a fairly subservient relationship with big tech and sort of do its bidding. And they clearly see that as somehow in their interests and since they identify with the nation, the national interests. And I when when I say they're AI peeled, I mean they they don't make any bones about it. You know, before they were elected or just after, there's Keir Starmer, who's the Prime Minister. The other probably most relevant person is a guy called Peter Kyle. He's the Secretary of State for Science, Innovation and Technology. And you know, he was saying even before the election that it was important to relate to companies in Silicon Valley rather than the basis of sort of you know having a relationship between a government and a corporation, it was important to relate on a diplomatic basis. So essentially sort of treating Silicon Valley companies as a sort of peer to the British state. You know, he wanted to have that kind of relationship. And anyway, since the election, they've just Google issued a report a few months before the election saying, okay, UK, you don't want to be left behind, you don't want to be a bunch of losers, what you need to do is dump copyright and build loads of data centers. And that's basically been their policy. They've just kind of sabotaged copyright, that's gone through, and they've got a big plan which includes building lots of data centers, and they've labeled data centers now as a thing called here critical national infrastructure, which you sort of get the idea, but what it basically means is it means a lot of things actually. But one of the things it means is you can't complain. You know, it's critical national infrastructure. So don't get funny ideas about trying to use planning legislation to object to the fact that you're going to have a giant data center built down the road that's going to deplete your already depleted aquifer or whatever.
SPEAKER_00Yeah, I feel like I've already seen a few stories of like local communities in the UK opposing data center projects that were planned near their communities and the national government just coming in and saying, it doesn't matter that you oppose it, it's getting built anyway, because this is the policy, right?
SPEAKER_01Absolutely. Yeah, yeah. And you know, and there's a few things going on there. The funny thing is, given the amount of rain there is here, that uh the UK is actually sort of water poor in some places already. And there are plans, for example, in the Southwest, you know, they've got a kind of contingency plan to import water from Norway on tankers or something crazy. And of course, that's tied into something that is underpinning a lot of this as well, which was basically it's Margaret Thatcher really, or her her as a personification of neoliberalism, because the water infrastructure is collapsing and has been in the hands of extractive profiteers for a couple of decades who've who've built nothing, stolen all the money, and we're left with l there isn't one waterway in the UK that isn't polluted by shit because these companies are being such a bad job. And that and that's obviously relevant for anyone considering the sort of material infrastructure data centers which place huge demands on water supply and huge demands on electricity. You know, our electricity grid is in a similarly sort of powerless state. So, you know, people are sort of rightfully and very immediately alarmed by the idea of these things being dumped into local areas, and they're quite right.
SPEAKER_00Yeah, and talking about the broader policy as well and like the consequences of this, obviously you talked about it. It was really startling to me, right, to see Google actually come out and be like, these are the policies the UK government needs to implement to be like, you know, a leader on AI or however they framed it. And then to see the government just turn around and basically do what they wanted, right? You know, to make it easier to build data centers and then to really go after, you know, the copyright protections that artists rely on to the degree that there has even been like a pushback campaign with some major musicians and other artists in the UK actually like directly opposing the efforts of this government to, you know, change copyright legislation to allow these AI companies to train on all this copyrighted material without having to like compensate or or worry about facing lawsuits from these creators.
SPEAKER_01Yeah, it is interesting. And these are things which which I think as our conversation sort of develops, these will all become obviously very relevant in the UK, but I think more broadly, but these are all sort of diagnostics, I would think of them as. You know, it's like the UK is very proud of its sort of creative cultural industries, you know, we've not got that much else going for it, really. So apart from the NHS, which is also big episode, we'll get to that, you know. But it is interesting because of that self-image of the UK and not just the self-image, but the actual empirical numbers on hard export earnings that come from that stuff, that it would normally be seen as kind of the crown jewels in a way. And to be sort of sacrificed like that, it's just very interesting, very notable. I mean, it should be it should be noted, it should be paid attention to, alongside other things. I mean, the that's what I mean about a sort of unified uh see, I wouldn't even say ideologically unified because there isn't really an ideology behind it, but sort of um stalinistically unified sort of government, that it is singularly focused on what it sees as its uh main goal, and everything else is literally to be trodden underfoot, including that stuff. I believe that the central thing, and this is relevant to the technopolitics of AI, I think, the central thing, which they say, you know, which Keir Starmer frequently, repetitively says, is that they are about growth. And they have identified, I think, that the AI industry equals growth for them. So that's it. Everything else can can basically go hang.
SPEAKER_00Yeah, unfortunately, we see a very similar thing playing out here in Canada where, you know, the new government is very clear that growth and the economy are the main priorities, and that also means, you know, incentivizing AI, giving these tech companies what they want. We've even seen that recently with the elimination of the digital services tax in order to keep them and, of course, Donald Trump happy. But I wanted to pivot then, you know, I'm I'm sure that the UK will still come up in the rest of our conversation, but I wanted to broaden it out, right? To start talking about some of these bigger concepts that you have been grappling with. And in a new book chapter, you relate what we're seeing to this concept of total mobilization. Can you explain what that is and how it plays into what we're seeing here with AI and, you know, these governments trying to do whatever it is that the companies want in order to encourage this industry?
SPEAKER_01Yeah, I mean the term total mobilization, I won't try and mangle the German. There's a book title by a guy called Ernst Junge. And you know, he is like a horrific character, but he's a really interesting character. He was a First World War veteran, most famous for his book, I think it's called Storm of Steel, his account of the First World War. But he he was a thinker, you know, he was a guy who, as I would understand it, you know, he saw the transition of the First World War from, as it were, conflict on a human scale to essentially a logistic, industrial, and even kinetic process. You know, he grasped sort of being right in the middle of it in a way. He grasped the idea that war had become something that was to do with energy and how much energy uh you know one could sort of enroll in energy in all forms, you know, mineral energy, steam energy, coal energy, and human energy, how much that could be focused and directed towards a single aim of of sort of total war. And there'd be no distinction between the civilian population and the military, there'd be no distinction between sort of the human and the machine as being part of a total assemblage, if you like, that was that was this war machine. That's one thing. But he kind of pursued that, you know, he pursued this idea with a very I mean I I do agree with people who believe that he is very influenced by Nietzsche. He has a nihilistic vision that we would now call accelerationism. You know, it's not really fundamentally different in that way, because uh his vision was to pursue this to its kind of ultimate conclusion as a form of breaking through to a different epoch. You know, he really saw this as a a turning point or a pivot point in the very mode of being, and uh definitely the very mode of civilization. He saw you can be sure his ideas of what civilization meant are like deeply problematic, right? But he saw this idea of civilization and breaking through to a new civilization, you know, with a uh a sort of new man, you know, and a new a new race, sort of powered by by this conflict form of technological accumulation of all available energies, and also through this idea of a complete transformation of society. So he it was a very yeah, we we'd now call it an acceleration, kind of extreme accelerationism. And it was explicitly, again, with relation to Nietzsche, powered by this a will to power, you know, a will to uh ultimate power over, over opponents, over planet, over being in general. Pretty crazy stuff. But having kind of stumbled on this idea, I felt that it was very resonant with what's going on because I feel that it's you know, and I'm not trying to impose it. I'm not trying to say that you know anybody else should take this idea up if they don't feel it fits. But what I do think we need to do is search for other ways of understanding what's going on, because it is irrational in a very, very deep sense. It's irrational, but not in a dispersed way. It's irrational with a sense of direction. I mean, even this stuff, I don't want to bring back things of the UK, but just thinking about this idea, or Canada particularly, but just thinking about this idea of uh identifying AI and economic growth and geopolitical strength, which are their kind of tick boxes. I mean that would be awful and brutal, but would have an element of rationality about it if this stuff actually delivered on that, which it doesn't, because the technology doesn't really work properly. It's just a bit crap, essentially, at a very large scale and in a very expensive way. And it isn't delivering massive productivity or anything else. So what I ask myself and I ask our listeners, right? What is actually going on? So one of the things I'm sort of offering as a way to frame this is this idea of token mobilization that what we're engaged in in here, most of us unwillingly, is a very authoritarian, very nihilistic mode of seeking a sort of ultimate transformation to a new order.
SPEAKER_00I want to dig in further, right, to several aspects of what you're talking about so that we can flesh this out for the listeners and understand it properly. And I feel like the first thing that I grabbed onto was this notion, and you mentioned it there, of being so focused on energy and turning things into energy. And I feel like when I think about the data centers and what is powering AI and what some of these tech CEOs have been saying, there's this total focus on increasing the amount of energy that we produce, then that we consume, also that we can enable this kind of AI-powered future that they have in mind for us. And that seems like, in a very direct sense, aligned with what you're talking about. And then we can talk about some of the bigger things that come out of that. But but I wonder what you make of like the actual kind of push for AI adoption, for creation for this industry in general and how that relates to the concept that you're talking about.
SPEAKER_01Yeah, yeah, absolutely. And and again, I'm really coming to grappling with these ideas by essentially doing what everyone else is doing, which is trying to understand what's going on in front of us. And well, in a way, in a lot of ways, what's being shoved down our throats, but what is completely unavoidable. I mean, I do find that this current wave of AI is interesting and provocative because it makes so many things so very visible. I mean, it's not like anything digital doesn't have a supply chain. You know, it's not like anything digital doesn't have toxic collateral effects and dependencies and so on and so forth, but we have been able to sort of roll along while more or less ignoring those things and you know, like maybe getting a bit worried about them in a different context. And AI just puts that stuff slap bang center, unavoidable, because of its, and this is where the kind of what one of the keys to growth is well, is this whole kind of obsession with scale, you know, that they must scale. It's like that in itself becomes a quite a cult-like core belief, you know, that the only way is to get bigger and bigger and bigger. And this is not a new thing. I mean, the expansion of empire absolutely depends on the idea of unlimited outward expansion. So that's the uh that's the way that the only way that model can sustain. So it's it's not like this is a new pattern, but it takes a very particular form, and it's taking a very extreme and extremely visible form. It's very visible to I would say like most people now that AI has these kind of material dependencies. And it's also very observable that the other stuff that these companies are doing. I mean, I I get I get this is probably some of what you're driving at. Whereas before, big data center providers, like AWS or Google or whatever, put a lot of time into performing sustainability, claiming green credentials and all this kind of stuff. And it's it's just so interesting that in the last year or so, they've not just abandoned those things, they've just kind of quite cheerfully abandoned them. They just say, well, you know, we can't do that anymore, and anyway, it's not important because this is what's important. And so goodbye, climate targets, yeah. And I would suggest that this is very could be very straightforwardly explained by saying our existing extractive economy just never went away, never even really had a second thought about continuing, never gave any credence to this idea of transitioning to anything else, and is just sort of blasting its way back through the gateway opened up by AI. And I I guess I I don't disagree with myself on that one, but I would think I I'm trying to reach for an understanding of this being a more coherent process, a more integrated process in some way. Not that there's a giant plan, but that there's something specific and possibly unique going on at this moment in time. That is, is those things and more.
SPEAKER_00Yeah. And I feel like as you're talking about that, the other thing that really that I really related to as I was reading what you were writing about this concept of total mobilization and how it fit into what we're seeing with AI in this moment is, of course, the geopolitical dimension of AI, right? Because we can talk about AI as this thing that is kind of driving economic objectives for these governments. You know, you were talking about how the British government is always talking about how growth is the main thing and it wants to achieve growth at any cost. But then when you're talking about total mobilization and what Junger is talking about, there's this notion that like that we are on the cusp of this new epoch and that there are all these governments kind of like lining up, or you know, all these forces kind of lining up to make sure that they're well positioned for this transition, regardless of whether or not AI is actually this new epoch, you know, generative AI in particular, these large language models, all this kind of stuff. Many of our governments have begun to accept that that is the case, or at least that is kind of my read on it. And that seems directly related to the type of thing that you're talking about, where you see all these governments kind of scrambling to get their piece of AI in the hopes that it positions them well for whatever this next stage of capitalism, our collective history, or whatever is going to actually be or or how it's going to look.
SPEAKER_01Yeah, absolutely. It's funny because you had the use of the term scramble there, it made me think of this old phrase, you know, the scramble for Africa, which was the previous sort of imperialist uh conflict from the sort of late 19th century, which I think is kind of relevant. You know, this is the scramble for AI, which coincidentally also involves a scramble for Africa because it's also a scramble for minerals, apart from anything else. I do feel that AI is in many ways actually a response to a general context of collapse. I think the optimizations of AI as it manifests socially and economically are actually pretty well suited to a sort of neoliberal status quo, because that is about very much about a sort of reductive, isolating essentialism, you know, in which things can be optimized. That's the market, that's Hayek and all that stuff. But yeah, I don't know. I think that that looks pretty broken right now. You know, it's it's run out of steam. It was always Was a load of rubbish really, but it allowed a lot of extraction on the way. But it's definitely run out of steam, and there's a kind of um pretty strong understanding that that is the case amongst these governments in a very empirical way. And it's all breaking. It's all pretty much broken. And maybe the Labour Party was a good place to start because, like the British government, pretty much everywhere else, they haven't got any answers to this. They haven't got any responses. There are some pretty obvious responses, obviously, which is redistribution of wealth and uh, you know, a turn to more ecological production and all this kind of stuff. And they're never going to do it. They are never, ever going to do that. So they have to do something. You know, and they they're sort of buying it. I mean, who knows how much they actually believe it. I've no idea. I don't know whether they even care, but they're buying into this idea big time. You know, one of the really, really concerning things to me is that, you know, I'm happy to sit down and snipe at AI and its kind of obvious failings and sort of ludicrous sort of performance in so many ways, which are also unfortunately very serious. But the other thing is I think it is quite good at being destructive. It's not very good at actually optimizing and actually sort of reliably predicting or generating something that can be trusted enough to be a functional part of society. But it is quite good at intervening in ways where essentially collateral damage doesn't matter. And you could think of that as anywhere from Doge to Gaza. You know, it's it's just it seems to play well in the destructive context. And that is to round off what I think you were just you know, potentially describing there. We're returning to the kind of context that Junge himself would have both recognized and relished. Essentially a war all against all, essentially based on not just nationalistic, actually, but sort of ethno-nationalistic terms.
SPEAKER_00And I wonder if it's worth, you know, before we pivot into talking about decomputing and what an alternative would look like, to double down a bit more and talk about those consequences, right? Because we have all of these governments kind of running to adopt AI, to adopt policies that are favorable to the tech industry, to try to get this investment, but also to claim that AI is going to make things more productive, that it's going to allow them to run governments more efficiently, that it will make their businesses more productive if they can just get them to adopt chatbots for whatever reason. But then on the flip side of that, there's this whole ignorance of the very real harms that we're already seeing from these technologies and a seeming desire, you know, not to recognize that they're even there. And to me, it even really stands out that like you can look back even two years ago, and it seemed like the discussion from a lot of our governments was like, okay, we need to adopt AI, but we need to do so safely. You know, we need to think about the ways that this could go wrong. And I even remember at the G7 meeting, I believe it was in Japan a couple of years ago, you know, one of the things that came out of it was like this AI safety thing. And of course, you know, we both know the issues with the AI safety framing. We've we've discussed them many times on the show. But then this year, coming out of the G7 was just like the statement on how we need to adopt AI for economic performance and, you know, all this kind of stuff, like the notion of safety, regardless of what that safety even meant, is like completely out the window. And it's just like AI for geopolitical positioning and economic expansion and any notion of the harms or the consequences of this, even as it feels like we have more and more stories reporting on the ways that chatbots are already affecting people, let alone the potential of AI being integrated into government systems and stuff, is already proliferating. So I wonder if you could talk a bit about that.
SPEAKER_01Yeah, I 100% agree with you. I mean, you know, you're you're a couple of episodes ago on your podcast was it was a great discussion about the harms that are being wreaked by things like character AI and people forming relationships with chatbots. I mean, it's it's really horrifying, actually. I mean, it's a whole domain of horror. And there's not only, you know, as was really clear from that episode, not only is there not any attempt to really constrain that, it's more like that scene as something to double down on, you know, something that can be intensified. My feeling and the reason I'm trying to sort of grasp for a let's say a you know a widescreen understanding of what's going on, a way to frame it in that sense, is because I am so concerned and frustrated, also with myself in this, about how it seems that we are going to be ever more continuously sort of dismayed and horrified by the effects of these kinds of techno-social systems being implemented in the world. And I really feel like that's sort of playing catch-up in a way. You know, it's how conscious this is, I I'm not sure in that would depend on different people at different points, different decision makers. With some is very clear, you know, you've got to listen to that idiot from Andreessen, who is it? Mark Andreessen or people like that, you know, delusional, deranged, hubristic, narcissistic, deeply unwell, but also unfortunately very powerful people. They don't make any secret about where they're coming from. But they're just figureheads in a way. We're talking about a systemic transformation here. There are many people involved who may not think of themselves as consciously engaged in this process. And yet I would say there is a clear, broad direction here, and if we don't get ahead of it, then we'll be in a very poor position to do anything about it. And I think it's very hard at the moment, I would say, to underestimate, and we don't need to, right? We are living through a genocide, right? So we don't need to underestimate the consequences of world changes at the moment and the things that are prepared to be tolerated if they're seen as somehow in the interest of those who are willing to, I don't know, you know, who whose interest is in maybe not just sustaining the status quo, but in further concentrating power, wealth, and control. And powered, I think a lot of the time by fear as well. You know, there's this kind of fear that they're they're going to get caught out. You know, there's a kind of fear they're going to get held to account. There's a fear of previous social movements which have been in the other direction. I'm thinking here, like of the the late 60s, you know, where where a lot of the things that we now consider to be rights we've become accustomed to, not that they are in fully in practice, but the at least conceptually, you know, racial justice or gender equality, all of these kind of things, you know, took another step forward. This is seems to be to this systemic transformation, its other, the thing that drives it, the thing that it wishes to erase, not to just to erase, but to completely roll back. And there doesn't seem to be any even any performative gesture towards, well, shouldn't we be nice about this or shouldn't we be ethical about this? Or let's not go too far. I think that's completely abandoned. And that to me is consistent with the idea that the the emergent drive is of this form, the kind of thing I'm trying to describe with this total mobilization. It is fuller, it is unabashed, it is really quite extreme. And we should consider what the sort of valid modes of response are, you know, in a way that tries to actually counter it with something not equivalent, but something that has a sufficient compulsion of its own, because simply trying to tread water is just not happening.
SPEAKER_00And I think that sets us up really well to talk about decomputing, right? If we're thinking about what some kind of alternative, some way to push back on this very strong effort to adopt AI, to proliferate it through society regardless of the consequences, to set it up as this thing that is this massive change, right? That nothing is going to be the same because now we have AI. Really great kind of marketing slogan for the companies and a way to sell their product to us, right? And so you have been writing about this concept of decomputing for a while. And I am really struck by it and find it really fascinating and think it's something that we need to be thinking much more about. So for someone who's never encountered this concept before, how would you describe what decomputing is and what it seeks to achieve?
SPEAKER_01It's an idea in development, or you know, it's or it's an open idea, really. That's maybe a much more positive way to put it. You know, I'm I'm I'd be really interested in this idea being sort of taken up and developed. Where it's coming from for me is is on a couple of levels. One is very straightforward and something that we would very immediately share. You know, it's it's looking at the you know, looking at the very material and immediate manifestation of a hyperscale data center and what that is doing, what that is not just representing, but what that's putting into operation and trying to address that in a way that's that doesn't split off one or other part of the consequences of that. So what I'm trying to say is a hyperscale data center is both a very physical and energetic, yeah, as in sucks up energy, concrete in large many cases, object. It's a real thing, it exists in the material world and makes puts a strain on our broad conditions for existence in a way that definitely does make things worse. But it is also a platform for things that make our life conditions worse in other ways. Without hyperscale data centers, it would be very difficult to have the forms of precaritization that we currently have. You know, the world of delivery, the world of Uber wouldn't work without data centers. So I'm trying to construct a response to these things as it that are delivered together. You know, it's something that's putting a strain on our literally our ability to continue living across a number of axes, if you like. Both the physical conditions of existence and the social relations and the productive relations of existence. AI specifically and hyperscale data centers and what they enable, what they platform, is both at the same time. And so, at the very least, decomputing is saying, stop, essentially, saying enough, you know, that these things are foundationally toxic in what they are and what they deliver, and therefore we should be pushing back and saying, no, thanks very much, enough of that. That's trying to see in a way why, and this is maybe relevant to a lot of the things we discussed, why do these things have the power they do have? Why is it that AI, despite its rather chonky nature, you know, it's kind of rather uh sketchy technology and the advocates are quite lacking in credibility. They're like very, very poor salesmen or something or other. Why, despite all these obvious factors, is it, as you were saying earlier on, you know, is it winning over some of the most powerful people on the planet? Why is it winning over corporate commitments, you know, to the nth degree? What else is going on here? And I think what to say about that really is the summary slogan you put there about them saying, Oh, this is going to change the world. My my reading would be because the world was was already not what we thought what we were led to believe in the first place. You know, it's like the reason why these things have such power is because the systems that we are already essentially subject to are already quite a lot like that. It's not simply the hyperscale data center and the AI algorithm. It's those things are riding on the wave of forms of institutionalization, forms of centralization, forms of political economy that are already delivering the harms that we're talking about, they're just being intensified by AI or delivering the climate damage that's already being intensified by AI. One of the things I try to pick out when talking about decomputing is the idea of scale, because that is a kind of skeleton key for a lot of these harms. It's a thing the industry itself fetishizes, it is actually quite an important part of getting any of these junky machines to work properly, and it is also a complete corollary for the general idea of growth. It's a complete corollary for the idea of growth at any cost without boundary, never mind what happens to people and planet. I guess decomputing is trying to be the other to scaling. It's trying to say whatever it is that leads us to believe that that is the way forward, whether however we're structuring institutions and technology, then it'd be better to do something else.
SPEAKER_00And I appreciate that framing of it as well, right? Because I really feel like looking at the hyperscale data center and looking at these tools also forces us to consider why does this exist in the first place? Why do we need this kind of infrastructure? How is this technology actually working? Why is the hyperscale data center needed? It is needed because we have this system of mass data collection, or we could call it mass surveillance on the public, this desire to create these forms of algorithmic management and control that require all of this computation in the first place. There are all of these tools and all of these forms of technology that have been built up now over the course of a couple of decades, a few decades, that seem to all be kind of coming together in this moment as the tech companies, as you were saying before, kind of push for this acceleration of all of it, right? Under the guise of generative AI and, you know, whatever new future is kind of coming into being here. And thinking about decomputing is not only just like, oh, how can we use less computing power or not buy a new laptop as often or something like that, but is looking much more structurally at the system that has been set up and the way that it has significant consequences for the way that we all collectively live.
SPEAKER_01Yeah, yeah, absolutely. Summed it up really nicely. You know, we we are social beings and uh societies are socio-technical or something like that, you know, or rather, the term I prefer is technopolitics. You know, if we're looking at how we want our worlds to be arranged and how we want to think about our power distribution and our sense of agency and our sense of collectivity or not, then we look at our societies and and our technical structures broadly. Now, technical structures does mean technologies, but it also means the institutions, it means the procedures, the laws, the protocols, and everything else. It's the machine of society that we're really looking at here. And this machine is not only broken, but clearly extremely vulnerable. I thought Doge was really a great example that way. We were all looking at this feeling initially flabbergasted. How does a bunch of repellent nerds with six laptops in each rucksack who were ushered in the back door, you know, under the aegis of executive power, so quickly manage to bring such a high level of destruction to, if not venerable, then at least vast, you know, institutional and governmental structures? And I think pretty clearly it's because of the way those structures already were arranged, you know. It's like this is a neoliberalism, as it wasn't only a process of marketization, but some other process of quite deep automatization. The market mechanism of all automated optimization was extended to all forms of social and governmental relations in a way that made it very ripe, I think, for somebody to sort of close that final loop and take it over and or destroy large chunks of it very, very quickly. So, yeah, absolutely. I'm just agreeing in a very long-winded way that what we're looking at here. And I think that works. Well, I think it this is uh an idea that works for me as a handy idea, because it works at different levels. You know, you can take this as a reading of the broader collapsonomics of society and how we're going to deal with it, but you can also take it to a local context. I mean, I work in higher education, so it fits there. It's like, okay, what is going on with the sort of Californian wildfire that's reeking its way across higher education so quickly with the introduction of ChatGPT into general distribution? Why did that, why was that able to cause such internal collapse in our educational systems? Well, I would say it's got a lot to do with the way that they were already constructed as essentially metricized exercises in data gathering and performative, measurable actions in substituting for relationships and natural learning, they're already like that. So along comes this very specific but relatively modest technical development in a lot of ways, and boom, you know, it causes it like it's like an explosion in a powder-filled room. But more positively, you know, decomputing being a stance on that. It's saying, well, okay, if we want to, like I say in my case, reclaim the idea of critical thought, reclaim the idea of learning, reclaim the idea of education more generally, you know, we want to construct, reconstruct, or reclaim forms of education that are other to this, that are not vulnerable in this way, that are not constructed in this way from the bottom up, not just ban AI or something like that. But really, yeah, I guess this is the thing. In every situation where AI is presented as a solution, in every situation where AI is able to wreak harm, because whoever it is in the decision-making role decides that that is the case, it is the case that we should re-examine what is going on there in the first place.
SPEAKER_00Yeah, because I guess in that sense, it's not just looking at the technology, right, and how the technology is being rolled out, but the broader structures that not just encourage the creation and the rollout of a particular kind of technology, but these broader policy decisions that have been made over many years that on the one hand, yes, lead to and not just policy decisions, right? But you know, decisions by government, decisions by corporations, you know, the broader way that our society operates, that on the one hand, yes, leads to the creation and the deployment of certain technologies, but on the other hand, leads to decisions that, as you're saying, you know, erode the education system, make it about achieving a certain grade on a test and passing certain metrics rather than actually engaging in critical thought and learning and all these sorts of things. And then we see similar dynamics in so many other aspects of society where you have had this kind of breakdown and this kind of hollowing out because of, you know, we could call it these neoliberal processes, you know, kind of broader pressures of capitalism on a social safety net and on aspects of society that don't turn a profit or that are not profitable enough. Like it seems like it's not just a concept that is focusing on technology, but but looks so much broader than that. And probably that also comes out of some of the other inspirations that you had that kind of come into this idea, right?
SPEAKER_01Yes, absolutely. Yeah. I mean, with the decomputing, partly for the sake of my own sanity, trying to boil it down to a couple of ideas I can actually remember from day to day. One of the ideas, which is not a catchy phrase, but I use de-automatization, and that's meant to address the stuff you just talked about, which for me speaks to you know the fact that, again, you know, the reason why we have such a crisis about uh losing control over things to AI is because we're already fundamentally stripped of agency or the capacity to think differently about what it is we're doing. So de-automatization, not having such technologised relations, particularly decision-making, and also degrowth. You know, for me, trying to think through the idea of technological transformations that are also addressing the climate crisis, that are also addressing labour justice, if you like, and the the broad range of the harms I would describe as sort of infrastructural intersectionalities, you know, this idea that the just a data center, you know, it sort of messes with your it causes power cuts in the electricity grid and it also leads to you having a crappy job where you get exploited by an algorithm. This is a form of intersectionality, and there are many forms of degrowth, but the one that the ideas that I cleave to, which are the more transformational forms of degrowth, are doing the inverse of that. They're saying, okay, degrowth is not simply about re-examining our fetish for GDP, but we're talking about how would we live otherwise? How can we make social arrangements for otherwise? When I'm talking about this idea of decomputing, it's not that I'm trying to invent some incredible theory out of nowhere. I'm trying to assemble things in that that I might find, that I think we might find useful in trying to push back. So the idea of degrowth being one of them. But another area which I think you might have been alluding to, but anyway, I'll bring it up, which is what I find really handy, is the the ideas of Ivan Illich from the 1970s. And he had a particularly key text from my point of view called Tools for Conviviality. So the first thing about that is when he talks about tools, he explicitly is not talking about any technology or even what it sounds like, screwdrivers. And he does mention that, you know, he says tools, screwdrivers, bicycles, but tools are also the institutions that produce, that produce even abstract things like what we call education or what we call knowledge. You know, tools for him, you know, cross all those boundaries, and then convivial, which is talking about sort of quite a lot to do with relations and quite a lot to do with livable scale. So, and again, that's why conviviality or tools for conviviality, Ilich's ideas in general are a good you know, starting will give us some traction, even if we're first thinking about AI and its harms, if you like, is thinking about how firstly that scale itself and the effect that has on us is important because Ilich's reading was, well, when when stuff scales up too much, we have to start adapting to the machine, if you like, rather than the machine being something that we can use for our own creative and autonomous purposes, machine being in this broader sense. So I think ideas around conviviality are give us a heuristic for how we might decide to do things. But obviously, the first thing and most fundamental thing, and I think this is really just like a kind of core part, you know, you like the idea of computing, you don't like the idea of computing fine. You know, you don't believe in the uh the darkness of total mobilization. But one thing we could at least maybe all agree on is the fact that technology should be subject to social sanction, essentially. You know, there should be some decisions about what society as a whole, what people in general, but particularly people who are already most disadvantaged, are going to be affected by the introduction of large-scale technologies. There should be some way of collectively influencing that, if not even having an absolute say in it. And this idea is sounds weird now, you know, but it doesn't. It's not weird. It wasn't weird back in the 1970s. It was very common to talk about how society should have a say in what technologies uh unleash on us. We're in a in a time, as you were saying earlier on, you know, we've we've already been through this stuff with social media, we've already seen how the internal logics of these large-scale and unbridled technological and pervasive in so-called innovations can lead to incredibly multidimensional harms. We've done that, and now we're doing it again. Great, even more intensely on an even greater scale. And one of the fundamental reasons is we don't have a say. So uh the conviviality, in order to pursue it in any detail, implies grasping for, developing, building for some form of control over the introduction of technology. That can also be at a larger scale and also be at a local scale, just in my own context again. You know, it's like, I mean, I'm a poor person. By the way, academia and its so-called critical faculties have just completely rolled over and gone, oh yeah, it's inevitable, isn't it? This AI stuff, we better embrace it and change everything. All that this uh social sanction means is that as people who know what we're up to, students and staff, we should look at a technology, carefully examine it on a number of dimensions, and say, Do we really want this?
SPEAKER_00It's absolutely wild to me that that is like not the case, right? That we have kind of ceded this to this industry, to these people who I would argue, and I think many people would agree, are kind of so distant from the kind of usual experience of most people in society and have become more so over the years as they become wealthier and more powerful, and are also kind of seized by these ideologies that are so distant from how most people think about the world and what a better future should look like and all these sorts of things. And somehow those are the people not only developing the technologies, but deploying them into the world and saying, oh, if you don't like them, too bad, because this is what we figure is the technology that you should all be using. Like just that fundamental notion that these are the people who should be making these decisions instead of us collectively. And actually, like, whether that is us all kind of getting together or having particular groups that can assess technology, or even like just expecting that our governments would do the very basic assessment of new technologies before they're rolling out, instead of always kind of being on the reactive kind of back foot when it comes to how we're going to respond to these things. And then, of course, not really wanting to do very much unless they threaten economic growth or or whatnot is just such a broken and flawed way of pursuing any of this. And so I wanted to ask you, I think you were starting to get to it at the end of that response. If we're we were thinking about like implementing some sort of ethos of decomputing into how we manage technology right now, how would you see that as being achieved? Is this something that is done through policy mechanisms and government? Is it done through collective action and mobilization? Is it done in many different parts of society? Is it an individual kind of thing? How would you kind of conceive of that?
SPEAKER_01I wouldn't close, you know, close off different levels of possibility at the same time. I think it's the unfortunate case that we don't even we don't even have regulatory capture right now. We have essentially state capture. And as we talked about from the very beginning, you know, the all of the entities in society with the existing concentrations of power seem to be sort of swinging behind the same thing, really. Um and even worse, are sort of abandoning a lot of their tokenistic claims to, I don't know, diversity, equality, inclusion, whatever, uh, and just generally kind of getting weaponized. So I wouldn't concentrate my efforts on trying to persuade them, I think, at this stage. On the other hand, on a positive note, I would say that it's possible and legitimate for uh for us at any level, like for me, you know, in not even in my university, but in my department or wherever it is, or any teacher in a school or any worker in a in any kind of context to say, okay, this particular AI for this particular purpose being introduced into my life or or into my healthcare system or into my children's school, whatever it is, is is not uh acceptable actually. And that does imply a level of self-organization, actually, some doing something about it. And I think that is harder to sell to people as a compelling vision because we don't have the experience of it now. You know, we've had sort of 50 years of neoliberalism, you know, that have removed most people's experience of collective empowerment, have removed most people's actual experience of of agency, really. You know, it it's it is transformative, or it can be transformative. You have the experience of actually being able to say no, I think, which is a good start with this refusal, and to to start to act on what they honestly think would be a better option in their situation. But we mostly don't have the experience of that. And I think it actually does start, it does start locally. It starts with people, even individually, but mainly collectively, getting together in their particular context and just saying no to start with. I think this is a there's a very good argument for the power of refusal and the sort of the pleasure of refusal in this case. You know, we we should be a little bit possibly try to be a little bit punk about this, or definitely a little bit solar punk about it, you know. And this is maybe the other aspect of it that, you know, we need a broader vision, but I'll come back to that in a second. I think the I think unfortunately, and this might just be where I'm coming from, the pace and ferocity with which these things are taking hold means that, yes, absolutely, things like, let's say, unions are incredibly relevant, even though the union structure itself is a bit like one of those one of those tools that has taken a certain form, you know, which is so I'm not really talking about a union, I'm talking about a union branch or a group of colleagues in the union or whatever. That is an essential part of this resistance process, or a community group, or whoever, these are essential parts of it. But I think the ferocity with which stuff is changing is very, very brutal and clearly authoritarian at the very least, and constantly edging further and further in the in the direction of the far right, essentially, politics and uh fascistic politics. I know this I'm probably gonna make myself sound like really extreme here in some ways, but I honestly think that it's important at this stage, again, to get ahead of it, think in terms of resistance. In the UK, for example, that this government, I honestly believe their absolutely fruitless attempts to persuade people that they are the party of growth and it's gonna deliver well-being for everybody are gonna have absolutely no influence on the massive anger, disillusionment, and sense of betrayal that really mirrors what has happened in America almost one-to-one. We've already had pogroms where people have tried to burn down hotels for immigrants. You know, we're not talking theoretically here. If this is a stance taken on by institutions, then I think we have to think very seriously about how to organize in ways that involve an element of self-defense and an element of parallel structures. Learning the lessons of the past is really important throughout this form of self-organization. But to swing back from this, the most important thing, and again the thing that's been sort of limited, eliminated by all these years of crushing and deadening neoliberalism, is just a sense of a better world. I think this was much more pervasive in the 1970s. For a number of reasons, people had a sense that life could be different, life could be better. Or let's say go back to the UK, where the things that are currently being destroyed, that are the things that, however grumpy we might feel about, you know, the state of the UK, we say, well, at least we've got the NHS, you know, or at least we've got the education system, you know, these are the things that are being dismantled and crushed and and thrown into the uh AI, you know, the AI shredder, essentially. These came out of a feeling after the Second World War where it wasn't just that people had gone to war and fall fascism and, you know, were trying to prevent London from being bombed and all this kind of stuff. It was a shared sense that people weren't going back. They weren't going to go back to the 1930s. And I think we should have the same sense. It's not that we're just opposing AI. We don't want to go back to that status quo that came just before AI, because that was also crap. You know, what we really need is a vision of, and I'm not saying it's one vision or a unified vision, but we needed a sense that life could be transformationally better. You know, that's also really vital just to sustain our our kind of collective existence at the moment. So I if I think about something like the idea of decomputing, yes, it is about addressing a technology, it is about transforming our productive relationships, and it is about taking technology back under control. But it's doing that as well within this broader framing of a revived belief in the idea of a of a better life for everyone. And I'm very keen to find out from others, you know, what their line of thinking and work is towards that, because that seems to me a massively missing part of the jigsaw right now.
SPEAKER_00I couldn't agree with you more, right? And and I think that has been part of the reason that the tech industry has been successful at kind of capturing our minds, because they give us some vision of what a future could be or is supposed to be, you know, regardless of whether it's one that is actually better or is going to work out for most people, but it's a future nonetheless that says that things can get better. Whereas it feels like, you know, we've kind of lost that in the political sense, right? This I this idea that we can actually make things better through politics rather than just through technology being created by Silicon Valley companies or something like that. And I think my final question, you know, to build off of what you were saying there, is whether you see any examples of movements toward decomputing or movements that are that are making progress. Like I'll speak for myself. You know, you were talking about hyperscale data centers earlier. I've been very encouraged to see the movements and the organizing around data centers and kind of the visibility of those infrastructures and how that has been a way into thinking about these technologies for people in a way that maybe was not so apparent before they became so visible and people started paying so much attention to them. So I wonder on your end, you know, when you think about decomputing, when you think about these efforts to question the technologies that are being rolled out and these broader systems within which they exist, do you see any hopeful examples there that come to mind for you?
SPEAKER_01There are movements around computing itself, actually, which are surprisingly positive. You know, there's uh they come under a variety of names. What they all express to me is the sense of, you know, which I identify with, because I'm in the computing department, you know, people who actually work directly with these technologies who recognize that. Now, I'm not talking about AI here because I think the current form of AI that we have at the moment is essentially pretty useless for for social benefit. But computing more broadly, I don't, you know, as has printed views. I'm still a bit of a fan of the internet despite the state it's in. And I think these could all be important parts. So so movements like perma computing, frugal computing, computing within limits, these are all identifiable clusters of people in technology, but they're they're to me, in a way, receptors. You know, they're they're waiting to connect with, let's say, other social movements or other initiatives that are kind of coming at it from the other end. You know, I would say, for example, I mean, this isn't the this isn't the optimistic example, but I'm just something that's top of mind for me at the moment is um the way algorithms are being deployed to sort of further punish people for claiming welfare benefits, particularly people who are disabled. I mean, yeah, it's it's a very broad tendency and is intense in the UK as anywhere else. But actually, that the movement of people with disabilities, disabled people themselves have some elements of this already. You know, that technology is already integral to their lives and how well or not they are lived. They have a technopolitics and they also have an understanding of a sort of transformatory technopolitics. They've got some sense of, and also a hands-on sense, because they're often sort of hacking the technologies to make them better for themselves. They've got a sense of how they could, how life could be built better so that society wasn't constructing them as disabled, so that technologies weren't rendering them as disabled, but were doing the opposite. And I think that's that that's a good, you know, is it's a very real thing. Disabled movement is like a really good source for a lot of this stuff, but it's also a sort of paradigm for the rest of us. It's like as an idea. Now, the example that I find the most inspiring at the moment, I don't think it's really got that much to do with computing, but it does have the elements of what I think of when envisioned this, which is a thing called the call that they call the GKN collective, the GKN factory collective. And what that is is it's a factory in Italy, near France. They were making car axles, and they were bought by a British hedge fund, which tried to do what hedge funds do, which is shut them down and convert it all into cash. And they occupied the factory and refused to allow this to happen. Now it's all very well and totally understandable and sort of all power to them for doing that. But what they did then was, to me, it's kind of figurative of a kind of response around technology in general and AI in particular. They didn't just do that. They said, okay, we need to be productive in some way because we've got to sustain ourselves. We've got we're not gonna do this just as the workers left behind. We have to do this with the community. So the collective structures they formed were with their local community, so it's worker and community. And they said, okay, you know, if we're gonna do anything, it has to be in the direction of a better world. We've got a just transition, is an idea that they subscribe to. So they said, okay, we need to make stuff, we want to make stuff, we we're skilled to make stuff, we don't want to make car axles anymore. What what are we going to do? And so they and they they even collaborated with the local university to sort of develop plans about what they could do productively. What they're doing at the moment is making cargo bikes, which is great. And they're also recycling solar panels, which is also great. And I think their own plans, I think, to make PUV cells of their own. And even more than this, they understand what they are doing within a much broader context. They understand what they're doing as of this idea of a just transition, as the idea of working towards uh a balance with the planet, working towards um equality and justice for themselves and their communities, and working in a way that is a is pushing back. I mean, there's the slogan they organize under is insorgiamo, which means we rise up, and that is a partisan slogan, and that's a very contentious thing in Italy because it's a fascist government and they're doing everything they can to erase the memory of the partisan. So they're saying, you know, no, we are part of this movement for a real world that is not new, it's been here, uh, and it's fought for its existence before, and here we are fighting for our existence but doing something very constructive and positive. And I just want to connect back to your last question in a way, which I guess, you know, the thing that that generative AR has again made very visible in some very specific ways, going back to the beginning conversation about things like copyright and whatever, also all the the data labor that a lot of us are now aware of that underpins all this stuff, is that there is nothing in these things that doesn't come from us in the first place. You know, there's no value, no actual value or or anything constructive or positive or creative in this at all that didn't originally start with us. It's been extracted, it's being um divided, it's being reduced and turned against us. But at the end of the day, actually, we already make this world every day, and there isn't any reason why we couldn't remake it. And that collective are doing it under duress, but they they're doing it. And I think the the onus really is on the rest of us to take that as an example.
SPEAKER_00I think that's a great place to leave it. You know, it leaves us with a great example, but also kind of brings this all together for us. It's been fantastic to learn more about this concept and how you are conceiving of the moment that we are in. Dan, thanks so much for taking the time. It's always a pleasure to chat with you.
SPEAKER_01Yeah, likewise.
SPEAKER_00Dan McQuillen is a lecturer at Goldsmiths University and the author of Resisting AI. Tech Won't Save Us is made in partnership with The Nation magazine and is hosted by me, Paris Marks. Production is by Kyla Hewson. Tech Won't Save Us relies on the support of listeners like you to keep providing critical perspectives on the tech industry. You can join hundreds of other supporters by going to patreon.com slash tech won't save us making a pledge of your own. Thanks for listening and make sure to come back next week.