Alix: Hey there, this is Alix Dunn. Welcome back to Computer Says Maybe, where I talk to the people pushing back on tech power and fighting for something better. Our last few episodes have focused on the growing influence of corporate tech power in US politics and community pushback, from local fights to broader efforts to build political power.

But these questions obviously extend far beyond the US. So for the next two episodes, we're partnering with Luminate to look at how communities around the globe are confronting the consequences of the tech industry's promises of progress and building alternatives rooted in justice and real public benefit.

We're starting in Europe, where the push for so-called technological sovereignty is gaining momentum as the EU looks to reduce its reliance on US and Chinese tech companies and strengthen its own position. But in its effort to compete, is Europe just adopting the same any growth at any cost mindset that's driven so much of the industry's most harmful behavior?

Today, I'm joined by Audrey Gaucheron, Lu Valgreen, and Max Schulz, who work at the intersection of data, climate, and AI across Europe. We're talking about the costs of tech sovereignty, the influence of accelerationism on European policy, and what a more sustainable vision for European leadership might look like.

Let's get into it. Audrey, you wanna go first and say a little bit about how you come to this conversation, your current role, kind of what SOMO is up to in this broad area?

Audrey: I'm Audrey Cockburn. I'm the executive director of the Center for Research on Multinational Corporations, and we're based in Amsterdam.

And it's a corporate power organization, so we look fundamentally at what has given multinational companies the excessive levels of power that they have. So the companies themselves, but the structures within which they're embedded, which in our view not only empower them, but allow them to dramatically increase their power over time.

And by looking at the situation that way, I used to work at Amnesty, and I worked on these issues from a human rights perspective when looking at the symptoms of big tech power. But over time, and in my current role, it becomes increasingly clear if you don't look at the structural element, you cannot deal with the symptomatic element of big tech power.

So that's where we sit at the moment.

Alix: Lou, do you wanna talk a little bit about how you and Data for Good come to this conversation?

Lou Welgryn: I'm Lou Vallgreen. I'm the secretary general of Data for Good. We're a French NGO, and we have several ways of acting. Our main objective is how we do tech for the social interest.

So the first point is really putting it to service of NGOs that are fighting for democracy, social justice, protecting the environment very concretely. For example, it means tracking big fishing boats, industrial fishing boats, and knowing when they're fishing in protected areas, enabling French citizens to know the quality and the pollution of the water they drink, stuff like that.

But it's also about fighting the concentration of power against big tech actors, and I think it's mainly the reason we're here today because in France we're doing a lot of lobbying against data center expansion in the name of sovereignty and pinpointing everything that it means in terms of real sovereignty, environmental impacts, social justice.

And for instance, in France we already have two ongoing litigations against dirty data centers with local groups.

Alix: Max, do you wanna say a little bit about how you come to this conversation on EU sovereignty and AI?

Max: I'm Max Schulze. I'm one of the co-founders of, uh, European think tank, Leitmotiv.

Personally, I've also been part of the first sovereignty, uh, debate that was, I don't know how many years ago, that ended up in Gaia-X, so this feels like a very recurring theme. But I do think there's a difference between, um, how sovereignty's hijacked versus how sovereignty is pushed by society. I think there is a principle desire to have my data processed in, in my country and that I get somehow a benefit out of it.

I think that's how I would say society is pushing for sovereignty. How it's then hijacked in policy conversation that it's not actually sovereign. But I do think the desire is there to do it right.

Alix: Okay, let's, like, take a step back and I think dig into this word sovereignty. Do you wanna try and define what AI sovereignty means?

What this is attempting to conjure up when policymakers talk about this?

Audrey: Aside from the dictionary definition of sovereignty, which is not what most people are talking about, I think different actors are defining it very differently and leveraging it very differently. So you've got policymakers in Europe, many of whom are in an existential tailspin about the identity of Europe and where it fits in the world, and, "Oh my God, the Americans just bit us hard, and what do we do?"

And they wrap that up also with the competitiveness agenda. So they've got an economic, political, existential crisis, and AI sovereignty means something in that context, although I'm not entirely sure they're clear what it means to them. But if you talk to business, when they talk about European AI sovereignty, it's opportunity.

It's the opportunity for the injection of public capital, for public procurement contracts, and that has its own dimensions. But then you look at something like Eurostat. And Eurostat gets a political and economic ideology that is very much pushing forward the idea of Europe needs to reassert its place in the world.

And I think that part of the sovereignty debate worries me quite a bit. I read a McKinsey report last night, and they were talking about what sovereignty means for non-tech businesses who are thinking of adopting AI, and for them it means something totally different. So the, the phrase has become a bit of a catch-all, and people having conversations, and they're literally driving past each other.

They're using the word sovereignty to mean different things, mostly in their particular interests, and they're, the clear definition obviously doesn't exist.

Alix: Lu, when you hear AI sovereignty, I mean, one, am I right in that it is coming up loads in Europe, and also any additions to Audrey's definition or how you all approach it?

Lou Welgryn: Yeah. I think what's really interesting that we observe in France is that discussions are really focused on data centers. So they are really trying to make us believe that because we're gonna have lots of data centers in France, we're gonna be sovereign. And the reality is that 91% of data centers that are being built currently, uh, are for extra European operators, mainly Americans.

So just to give you some numbers, in France currently, we have 2.3 gigawatts installed for the moment. The biggest project that is being built is 1.4, so it's 2/3 of the size of every data center ever built in France. And currently, o- There are 18 gigawatts that are asking to be plugged to the electrical system.

This is the equivalent of one third of our nuclear power. So in France, we really have a plug baby plug strategy, which is done in the name of sovereignty, but it's not even building any kind of sovereignty. It's rather increasing the dependency to American tech companies today.

Audrey: Well, someone's also done a study on what's behind European data centers looking at a number of countries, and there's the US, or the builders of many of them are US companies.

And when they start off being European companies, they've been bought out by US companies. And behind those US companies are investors like BlackRock, State Street, Vanguard, all big US asset managers. And I think this is the fundamental incoherence of everything to do with European AI or tech sovereignty.

Europe believes before everything else in the free movement of capital. Europe protects the free movement of capital more than any other bloc or country. Even the US, which is a big protector of free movement of capital, protects US economic interests within the framework of that. Europe just protects anybody's capital, and that's such a founding principle of the European Union.

It's why there was so much Russian money and why it was so hard to deal with Russia at a certain point. Because for Europe, all money moving wherever it wants to go is almost a religious principle. So how can you talk about sovereignty when you're talking about Things that are owned by private actors, and it can be owned by one today and another tomorrow, and that goes for critical infrastructure.

Europe has seen critical energy infrastructure move out of European hands into US hands with literally no regulatory oversight, and the same can and will, and already has happened in relation to, for example, data centers.

Alix: Do you all think that sovereignty is a concept in a digital world that is this financially globalized?

Do you think it is a non-starter? Or, like, how do you engage with that discourse?

Audrey: I'm struggling to see the positives with it. Yeah. You can't achieve sovereignty in any meaningful sense. What you can achieve and leverage it for is a certain amount of safeguards. But if you continue to operate in a context where you're unwilling to put restrictions on corporate ownership and corporate capital, or sufficient restrictions, then you're doing what the European Union does best.

You're writing loads and loads of laws, many of them deeply irritating and hugely ineffectual when it comes to the actual impact that they're trying to control. Yeah. And companies can move out so far ahead of that in terms of-

Lou Welgryn: Yeah ...

Audrey: how they structure their operations, how they structure their subsidiaries, who's got what identity.

They will run rings around the European Union. So I think it's more about the leverage for having a different conversation about where we're going with AI because that is not happening. The pace- Mm ... it's literally, let's just throw up data centers all over the place. We must win the race on AI. That's a narrative big tech has promoted.

They have inculcated it into policy makers. Mm-hmm. They've used their communications channels to push it out around the world. We did work in multiple countries. The same narrative is everywhere, and it's all about speed. And what they want is, "Don't stop us. Don't stop to think. Just do, because your entire economic future depends on doing this."

If we can use sovereignty as a way to get into a conversation with, in Europe, about we need to stop doing this. You can't win. You can't really be sovereign. You allowed huge amounts of concentrated power to build up in the US big tech giants. Mm. You can't unwind that. We need to figure out what is it we're building all of this stuff for.

Some of it can definitely be good. There are positives. But there's also the risk that you throw up a whole b- bunch of data centers around Europe, and they need customers 'cause they're businesses. And when they need customers, their customers will be the big tech giants, and about how much social media slop of meme factories and rubbish will just be...

Because that generates money, which is the purpose of business. Mm. It's not to make a good or a product. It's to generate money for shareholders. So I think the sovereignty debate, can it be pivoted in any way to talk about the bigger model issue? Because if you don't do that, you're just gonna slap lots of plasters all over the place, and there'll be like a stack of European regulation, but it won't change very much.

Lou Welgryn: Yeah. There is a French, a researcher that's called Ophélie Quenault, which is specialized on technological geopolitics, and I really like the term that she's using. She's saying we should not talk about sovereignty. We should t- talk ... be talking about choosing autonomy. So the idea is understand our dependencies because once again, the debate is all about data centers, whereas tech world depends on cables, on models, on data collection, on research, um, tools, et cetera.

So she's really forcing us to ask this question, what dependencies do we want to have? And, and then what are our, what are the, the usages that we wanna have, and prioritize them. And as you were saying, Audrey, the, the idea is really Why are we building all this AI first and all th- these infrastructures?

Do we really want our children to be learning at school th- and that the only things that they're learning to write is prompting ChatGPT and stuff like that? I don't really th- think so. But at the same time, they ha- there are some questions that need to, needs to be answered, because for example, in France, our health data are stored on Microsoft Azure servers.

So this is a problem. This is a problem when we know the Cloud Act, when we know what can happen to data, and when we know what Donald Trump wants to do to Europe currently. So I think the questions about sovereignty should also be a question about we have some limited resources, how do we preserve them?

We have a limited amount of water, of energy. Where do we want to use it? And there is a limited amount that can be dedicated to data centers, to tech, and then this limited amount, what do we want it to do? And I think the main question we should be asking is, who is benefiting today from this development?

It's only American companies, data center builders, et cetera, and in front of us, there are people always defending this, there's gonna be shared wealth for everybody. It's not true. I think this is the first narrative we need to unbuild is that it's actually only beneficial to a handful of companies.

Basically, today, the European strategy is completely AI first. Let's put AI everywhere, and we'll, we will look afterwards at the problem and deal with them today. I think we should have the opposite strategy, which includes precaution, which ex- include in what kind of society do we wanna live, and then we can answer what, which technology to develop to answer this society and to defend its value.

Audrey: I think what you're describing, that is sovereignty. Sovereignty is when you are having a genuine public debate and allowing that public debate to shape how Europe goes forward, and that's the one thing that's being foreclosed at the moment because this urgency, we must win the AI race, something catastrophic will happen.

I also think that there's another dimension, which if we did the things we should do now and pull back to a certain degree and ask questions, AI for what? What do we want it for? What do we not want it for? We also have to recogn- in a capitalist system, everything is interconnected. Pension funds are investing in AI.

There are businesses that are believing that this is the way for, of the future, and they've invested in certain ways. There's an awful lot of economic fallout from any of the things that we're talking about, and this, th- it's not specific to big tech. This w- this was a huge, I think, blind spot of the climate movement not to recognize the fallout on things like pensions and investment because If we don't deal with those hard realities, everything we say can be overwritten by a perfectly legitimate, if you do that, you are going to doom a couple of generations to recession.

But that, that's not necessarily what has to happen. What has to happen is a conversation about how we manage the risks of going in a different direction and having enough space to do that. For me, that's genuine sovereignty, or at least the first step to genuine European sovereignty in this area.

Max: I often use the word economic sovereignty because fundamentally it starts with, okay, I have this infrastructure in my backyard, what's my benefit to this?

And in the digital economy, all the costs are always localized, so I get the infrastructure, I get the energy consumption, I get the pollution, I get the water use in my town But usually all the upsides or all the value creation happens somewhere else because all the computation that's created by this infrastructure, all the services that are running there are traded on the internet, which is basically a global unrestricted marketplace with no tariffs, no borders, no taxes.

Where, in other words, I can put my credit card into a US service, pay in US dollars, but the actual infrastructure may be next to me. Um, so n- none of that money ever flows through to our jurisdictions, and I think that's the first fundamental question. How do we localize the benefit? If we accept the localized cost anyways, then the first reason is, like, how do we get the value creation back?

And I think that's a very important point. There's a strong misunderstanding from the old times. I build a factory in my town, obviously I'm gonna accept some cost from that, but I also get labor impact. I get maybe even a cultural impact of this factory, create some, some hub where everybody has their identity from.

It used to be very simple, and I think a lot of politicians think still a data center is like a factory, and I get all these benefits if I just accept the cost, and that's the first tooth to be pulled. And then the second one, and that's, I think, pretty... It's very evident in France, but also in Germany and in all European countries, is that the energy companies and the utilities actually know that with the push for energy efficiency and sustainability, energy consumption over time is gonna go down, and that threatens their business model of a lo- a lot of the power plants, especially nuclear power plants.

And so there is a very strong policy push also from the energy sector to accept data centers because they will pay for power, and paying for power means we can have power plants operate profitably for longer. Um, and that's, yeah, it's basically going back to the old days where we tried to finance our coal, gas, nuclear mega power plants that are way too expensive, and basically hurts the transition to renewables which actually cost the utilities money because they need to invest in, in, in new power plants.

They rather milk the old ones than to build new ones, and now somebody gave them a very easy way to do that by just saying, "Let's just build data centers. I don't care about anything, but they will pay for power."

Alix: Audrey, do you think it's possible to separate out economic sovereignty from these technical questions?

Or do you feel like structurally it's still just not possible given the current construct of how companies move money?

Audrey: If you wanna change where the benefits flow, and I agree w- with everything Max said But then you're organizing and structuring an economy because where the benefits flow and how things work is based on who's gonna make money and who's gonna make the most money and where the investment will go, which is why climate change has never been resolved, and it's why power providers will be very happy to see a new customer.

I think you'd have to interfere with the economy in a way Europe has never been willing to do, and sometimes that interference can be very negative, so I'm not necessarily saying that would be a good thing. But how would you go about, in, in the example you gave, Max, if you're buying something online and it, the sort of power and compute power that's generating whatever it is next door to you, but you're buying it from an American provider.

How would we even organize an economy to say that should be within Europe? Because the economy is organized by where investors put their money, where capital goes and flows to. I think more organization of the economy and more interference is a good thing, but you have to really think it through because there can be huge negative fallout from that.

But for me, as long as we're talking about the system that is pervasive in Europe and the US, the West, you are talking about a system where honestly it doesn't matter what the companies are making. They could be making T-shirts or oil. They could be making anything. That's not the point of the company.

The point of the company is to make money for shareholders, and it doesn't change that much if you haven't had an IPO and you've got venture capital because those guys want their money as well. They want their inve- a return on their investment. So when you've got that system, we're, we're talking about things that it sounds strange to say, but that's not so relevant to companies, what happens, the externalities, what happens, what the product is, whether it's on the moon or in the US.

It's what is the system that lets me maximize my return to shareholders, which is why a power provider can look at burning forests in France and see people dying because of the consequences of climate change, and the CEO will still think, "Yay, data centers. That's a solution to my maximize shareholder value, grow the business," whatever metric that CEO is judged by, which will always be economic.

As long as we're in that system, I don't know how you can make those changes, and if you did wanna make those changes, it takes us... In Europe it will take- the space to have a really good conversation with the public because there are fallouts and trade-offs to be made in changing some of the economics.

But if we don't do it now, what we're seeing will just continue. There's no amount of regulation Europe can do to stop what we're seeing at the moment. It can just make the worst of it not happen.

Max: One picture that I always have in my mind, because I'm also a citizen at the end of the day, right? We have created all these social networks, for example, and they don't function the way we had originally intended their function now.

And one simple thing that I would hope to have at some point is simply I want an Instagram for my town, um, because I don't really care what everybody else do. I just wanna connect with people around me. And I feel like that's more the sovereignty discussion that I think society's envisioning. Like, how do I connect with people again through digital tools?

How can they serve me rather than somebody else? How can we get rid of all the ads? And I feel like that's more the sentiment. To get there, to build on what Audrey said, I agree. You would have to interfere with an economy, but I would also say that the digital economy today is the one where we have done the least interference of all.

I think, if I think of US law, most of the laws are there to protect the freedom of that economy, like, to make sure there's nobody liable, nobody has to do anything. In the EU, we have created some vehicles like DSA, DMA, the usual, but we don't, of course, enforce them because the, the freedom of that market creates so much prosperity.

There's so much capital flowing into it now, and it's also a good indicator that it must be a very free and unregulated market because of how much money goes into it. I do think if we can separate finally two things from each other, right? So we have compute which is a commodity similar to oil, T-shirts, whatever the commodities.

Um, and we have services which are, let's say, software engineers creating a lot of value by gluing some technology together and then providing a service. At least on the compute side, I think we can all agree that this is, at this point, has become a complete commodity. And on that side, interfering with the market is something we have done in energy before.

We've unbundled, we've reorganized. We've said, "If you own a power plant, you can't also sell the electricity to consumers. You have to go through an exchange." And all these ideas are not new. And for the commodity side, I think, and especially again for the data center side, there is a lot more we can do. We can also talk about community-owned data centers, communities making the compute they need to power their services and forcing companies to use that local capacity.

We can talk about the state owning some of the compute capacity, as in France, all the power plants are owned by the state. I think like looking at that side first, because it's very fixable in that sense, it's very solvable, and then talking about the very complicated bit, which is the services, which is where also Eurostack and all these conversations get really complicated.

How do you replace an Instagram or how do you replace a, a whole Meta or Google or something like this? And we get hung up on how do we build a European search engine, which is a second order problem to like how do we get the compute in place to actually do this in a competitive way?

Audrey: To me, one of the biggest dimensions of this particular sector and this particular issue is the involvement of companies that also control social narratives.

They control whether Politicians get elected, and politicians know they have that power. They shape narratives. They shape our belief about what is possible. Aren't they going to deploy their vast resources on the communication narrative building side to scare the pants off the public about what that would actually mean in terms of the economy?

It doesn't have to be true, in fact.

Alix: And it is a, yeah, question for Max and Lu, like the extent to which the information environment within which this conversation is happening being- Yeah ... under the control of the very companies that benefit from a certain trajectory. What, how do we navigate that sort of elephant in the room?

I think it's a really good question.

Lou Welgryn: Okay. So to answer that, of course, I think the narrative is completely taken over by big tech companies, and one of the last example we had in the world was Jacob Coxon, the Mission resignation, where I really literally had journalists sending me questions like, "Are we all gonna die because of AI?"

They were so trapped in, and it was so massive, this narrative, that everybody was not knowing if they should believe it, what's behind, understanding all the marketing also behind it. So I think we're really trapped in these narratives. But I'd say the, the good side is that I think that since the second Trump's election, there has really been a shift in Europe about how we perceive technology.

We've really moved from some, from something that is a neutral tool that is not, where it's just a question about using it well or not, to something that's really becoming more and more political and shaping a vision of society. Uh, in France, we had a really, an example that really shifted and was really strong in the society.

It was the example of the international judge, Guillot. I don't know if you've heard about him, but basically he delivered a warrant against Benjamin Netanyahu And he was banned from all the American digital services. He could not pay with his credit card. It could-- He could not book a hotel. He had to travel with suitcases with cash in order to be able to pay where he was going.

So it's actually... This is the dependency that we have today regarding our t- our digital tools, and it's actually everywhere. So this kind of events have really shaped, contributed to shape the narrative on the other side and to understand what's at stake here. And I'd say the last thing to illustrate, I think that's positive thing, is that there's been-- there are some polls, so there have been polls in the US saying that people are 75% against data center.

In France, there was a poll recently that was showing that 83% of people want data centers to be the last element to have access to electricity in France. So it means that people are starting to understand what's behind it, and that it's actually jeopard-jeopardizing our sources. But it's true that we have a long way to go in order to really develop those narratives.

And I'd say the main problem is money. They have so many money in order to influence the debate that we need to use other things that are mobilizing civil society differently to fight those narratives.

Alix: The same thing happened to Francesca Albanese in terms of the role that she played in UN rapporteuring on Gaza.

The structural dependency on the US is so intense, and I think also maybe that this power was accumulated for such a long period of time before it was exercised in such a petty way. I think we're also seeing it being exercised and rulers that are willing to sort of demonstrate the dependency by using it for things that are actually quite, um, egregious and easy to tell stories about.

I think that is also probably creating some irritation, but now it's visible. Do we think that the fact that you're seeing these cases now, the fact that people may look at Visa and be like, "Oh-" That's actually a switch that the US could turn off for me. Do we think that public consciousness opens up political space to actually do some of the more dramatic structural stuff you guys are, are describing?

Max: I'm an in-infinite idealist, and we still have democracy. So I think though there's a scope issue for people working in digital policy, I think reminding people that we still have democracy and it's still working and we still have power can change anything, and it can also change the digital economy, can do all of that.

But we need to come together again and find our power. If you really look at the scale of, let's say, the top five tech companies in the world, they are de facto the size of states, right? They employ as many people as states. They have as much money as states. It's no surprise that they wield power because we gave so much power to economics and not just to pure military might.

And in the economic sense, they are very powerful states, and they wield it accordingly. I do just on, uh, on that, in that same bubble of thinking, to Audrey's point earlier, still businesses though, and if you see how much money they're now spending on compute, I think every shareholder, as you mentioned earlier, this is about shareholders, every shareholder is get-getting a little bit worried, why are you putting on so much risk and you're building all this compute?

Is this really necessary? And if there were an alternative for them sourcing compute at lower cost than doing it themselves From a business economics perspective, all the shareholders are probably like, "Actually, can you just buy it off the commodities market and not build your own farms? Just buy the milk from the milk market and not maintain the cows.

I know you can probably do it 10% better than everyone else, but it's not good enough for us, so just, just outsource the compute part." Because there they have the least IP. They don't, often don't own the data centers, they just lease them. They lease the land. They don't make the power themselves. They don't make the chips themselves.

They are now moving that way, but for a long time, at least for them, this was always a cost center, not a profit center in that sense. Um, and I think there's a way to push them that way.

Audrey: Europe waited too long, waited 15 years too long to deal with the concentration of power of the big tech companies and what they were doing, and Europe's own position.

Europe's gonna have to face a reckoning at some point. Either you float along behind all of this, or you do what they're doing at the moment, which is also throw a ton of public money at it. Didn't they say it was 200 billion that they were throwing at it? And then, somebody correct me if I'm wrong, but didn't Anthropic's IPO mention 500 billion?

And you think, what are we talking about here? Let me separate a few things. I think if companies wanted it the way that you described it, Max, they'd already be pushing policy in that direction, 'cause they'll push it and have the power to push it the way it suits them. So what they're doing, and the way things are structured, they're that way because it suits them.

Nobody has forced them into the position they're in. If they wanna transfer the costs to elsewhere, if they don't wanna deal with the costs, they will come up with a strategy that works for them, and I don't think the European Union's ever going to be in a position to do something that's beneficial to Europe, unless it's an ancillary effect of what's beneficial to them.

They'd be quite happy, for example, if public money is used to cover their costs. Business always is. Oh, great. But it will make sure that it's locked something in to protect itself. But the one place I feel that there's hope around democracy is actually all the protests against data centers, localized grassroots democracy.

It's so energizing, and we've actually moved our work. It's almost that we're supporting individual campaigns. But if you really wanna know who owns the data center that you're opposing, and how the arguments about how it's going to benefit you fall apart when you realize all the money's going to the US, we're giving you that data, and it's really energizing The kind of democracy I have not seen in a very long time, and I think we really need to make sure that insofar as we can, like we support that kind of grassroots activism because that, for me, is, uh, in quote marks, what

Max: will save us Double underline that because that's exactly my point.

In order to get them to change and to get rid of part of their business, we need to block some part of their business activity, and we are doing exactly the right thing. We're using democracy, we're using local activism to block data centers from being built, which means we're constraining them to build more compute.

And then if we can create an alternative that is maybe owned by us and leave them no choice but to either you use our compute that's community owned or you don't build anything at all, that's a valid way to push them that way. And I believe that can work if the policymakers, and I agree with you, understand how to support it in the right way, and that's not with throwing money at it.

Lou Welgryn: I, I completely agree that what's happening on data centers in Europe is super interesting, and it's actually really scaring data center builders. So in France they're talking a lot about acceptability. I saw that Microsoft recruited, I think something like a data center storyteller. Amazon is recruiting water storyteller.

So they understand that their narratives are at stake. And at European level, the, the network is becoming stronger and stronger. And what's super interesting is that it's coming from people living near to data centers, that for the first time in their lives, they want to act for something that's happening near them.

So I think this is super interesting. What's also super interesting is that people, they really-- it's not a not in my backyard movement. So yes, of course, there are people that just don't want a data center near their place. But what I really feel is that people are understanding what those infrastructures are for.

They understand that they're changing, uh, the way they access social services, that they are bringing surveillance to work, that they are having so many impacts in their daily lives, that they're not only talking about local issue, but a more systemic thing about why do we want AI for. And I think this is the main question in the end, because the infrastructure has to be structured in the way to answer defined needs.

And the questions of governance in the end that you were talking about, Max, for me, are really the center of this discussion, is who decides what we do with this technology, for what, and what are all the places where we don't want it, basically.

Audrey: The energy that you see in the grassroots movements around data centers, where have we seen that before and not for a long time?

There's a lot of energy around climate. Now, I've been in the climate movement for 30 years, and I think it's safe to say not a success in the end, no matter how much effort was put into it. And one of the biggest obstacles was that the industry, and we're talking a very powerful industry when you talk about the fossil fuel industry and the energy industry and, and the users, the big users.

But when we talk about big tech, we're talking about Hugely more powerful even than this industry. But it was able to combat every effort, every common sense thing, every civil society piece of activism. In the end, and I know I, I may sound like a broken record on this, but in the end, by persuading policymakers, and they didn't have to persuade too hard, "Do what those activists, those crazy people are asking, your economy will be destroyed.

You will have unemployment queues. People will be super angry. You will never get elected again." They told that story, and the reason it worked is because there was a significant part of it that was true. And as activists, we didn't wanna hear that. We kept talking about all the things that were value-driven m- things for us about the health of the planet, what should matter first.

But for a government, if somebody tells you genuinely, "You pull the plug on fossil fuel energy in the way that these activists are talking about, you try to just stop oil," and, look, I am speaking as someone who spent a long time in this, in this movement. If you don't deal with the fallout, and we don't talk about what are the structural economic issues that the companies are not wrong about, but come with our ideas for, "Yeah, that is a consequence.

Here's how to manage it." If we don't do that, I'm worried we're going to not learn the lessons of the climate movement, and that's 40 years of activism that has ended up with us seeing essentially goodbye to two degrees practically. We're in a terrible place on climate, and we know it's not gonna get any better.

What can we learn from the climate movement and do differently here when we understand that powerful companies will make arguments, and not all of them will be wrong, and some of them we have to grapple with the hard realities? And one of the things I really struggled with talking with colleagues, and I'm actually in a dialogue on alternatives to the way we structure our pensions, is the fact that European pension funds are investing in AI because they need growth.

We're getting older. More people are older. They need to grow their funds to pay their retirement. So are we telling people, "We're gonna pull the plug on things or reorganize things in such a way that, by the way, your pension fund for this generation and possibly the next will be s- like, significantly more difficult to pay out in any reasonable way"?

I think it's a legitimate question, and I think we can come with answers, but I think sometimes we're, we're so focused on where we are, we don't recognize the legitimate problems that our Policy or our policy prescriptions or our- Mm ... recommendations will have. And that for me is an important lesson of the climate movement.

Lou Welgryn: I completely agree with what you're saying, and I say the interesting thing here is that the difference with the climate movement is that for many people in Europe, it looks like something in the future if they have not experienced, uh, it for themselves, which is more and more happening, but it always seems less urgent than other stuff.

And I think that the difference here with the tech industry is that impacts are really tangible in a lot of people's life. At Data for Good, we have thousands of volunteers that work in the tech industry mostly, and they are being forced to use AI at work. They don't want to. They feel their work is becoming useless.

They want to become farmers, or they want to go to more manual stuff because they're losing the sense of what they were doing. The only positive thing about that is that we may have more and more people that can become aware and feel touched and concerned about what's happening. I say it's... or it's the problem, and at the same time, it could be if we manage the solution.

Can

Alix: I ask Audrey on this question about what we can learn from climate? It feels like there is a kind of a, a rhyme between degrowth as a frame and slowing down in AI. Basically, one of the only things I've seen work is, like, is this technology dangerous? Do we actually know what's happening? Should we actually be a little bit more cautious?

Cautious might mean go slower. Do we think that's... we've learned anything from some of those degrowth conversations, or do you guys see that parallel? And when you hear the slow AI discourse, is there an opportunity there to encourage cooler heads to prevail, get people to slow down, but also critically think?

Is it possible to do those two things?

Audrey: The fact that people are worried and the fact that slowing down in this particular context doesn't sound like the worst thing in the world to a lot of people does offer an opportunity. It sounds like the worst thing in the world to policymakers in most countries, though.

And how do you take the opport- how do we take the opportunity and look at some of the fundamental structural issues which don't just affect the tech sector and AI? I think we should be using the opportunity to talk about Degrowth, post-capitalism, slowing down. You don't always have to use those words 'cause they freak a lot of people out.

But I don't know that we are a big enough movement of people who want to use the opportunity to say, "If we slow down, can we also talk about some of the structural economic issues that need to change?" Because they would change, not just for this sector. However, yeah, it is one of the few times where s- the word slow is not going to put everybody off, so why not

why are we not trying to push that opening a bit wider?

Alix: Yeah, interesting.

Lou Welgryn: The, the growth narrative is also pushed by the tech companies themselves. Anthropic CEO wrote all this paper where he's asking for regulation, et cetera, but we know the story. Elon Musk did it three years ago. He was asking us to stop the development, and the

what was he doing in the meantime? He was developing his own generative AI. I think the thing is, follow the money. If you read between the lines of what he's, uh, proposing, Anthropic, when you read it at first, it sounds reasonable. He's asking for regulation. But what kind? What is his favorite kind of regulation?

It's actually an organism that's financed by the same people that are financing him. So it's the usual story of tech companies that are going to regulate themselves by themselves. How could we trust these people for regulating themselves when they're at the same things, that they're going to destroy humanity, and at the same time telling us that they're doing it for the greatest good?

So we should just Not trust them at all. And of course, we need to regulate... We hugely need to regulate, but we should not buy their discourses because they are really just making fun of us.

Alix: Yeah, no, I'm very much with you. I, I think that there's a breakthrough of the, this stuff where slow is publicly resonant now, even if the public doesn't necessarily understand the underlying things.

Also, this idea that super intelligence, whatever, is gonna save us all or it's gonna kill us all. I don't know. Have you all heard the expression from Ruha Benjamin that there are two sides of the same Bitcoin? Which is one of my, one of my favorites. Yeah, no. So I hear you, and I, and I don't think we should trust them as far as we can throw them, and I think them calling for regulation is just so ridiculous, and they're man babies, and we shouldn't listen to them or trust them.

But I do think that there's a, there's something here around this, 'cause Audrey, you said at the beginning of the conversation that basically there's no time to think, there's no time to move, and that basically policymakers are just being told to go. And I think that, like, if, if this moment of should we and where are we going, and do we, do we wanna go there?

It feels like for the first time that has broken through in public discourse globally because of some of the, the coverage, but maybe it's also just a, I don't know, it's a bad idea to get on the bandwagon of discourse that's coming from such toxic places.

Max: It's a recurring theme. Look, I think it all builds on top of each other.

First of all, AI companies have massive sunk costs, so to use Audrey's language, they have sunk capital. They want a return. They need to grow to get that return. There's no other way. And what they're asking is it, for permission to form a cartel. It's pretty obvious. So they can really control all the outcomes and ensure that their investors will get a return.

I think it's really simple. They're asking for carte blanche. I think on the degrowth thing, I actually think we've had peak growth. There's all these beautiful statistics in my home country, like the average German has 2.3 cars, uh, 1.4 houses. I think we are at peak consumption of what we can handle. And you, in, in private equity, you have this nice word called dry powder, meaning there's so much capital looking for a return, it doesn't know where to go.

And then you all of a sudden you open this digital economy, and every f- 10 years we have a new theme. So the first theme was like, can do social media, we can do ads. Then the next theme was smart cities, then there was Bitcoin. It's basically like saying to capital, "Just put the money in here, and we'll make it 10 X.

That's no problem." So it's attracting all this capital. This economic growth, that's pure capital growth, as Piketty from France, a famous economist, has proven, will never pay for the pension. So actually, doing social justice or redistributing money remains the number one thing we need to do because labor incomes contribute to pensions.

Labor incomes are not going up. Labor is less in the digital economy, and labor is potentially also impacted by AI And th- that's the real issue we need to solve. If we're really worried about our future and our pensions and our social system, the, the question is how to put labor back on top. Um, unfortunately, I think Piketty concluded that the only way to reset it was war, but I hope that's not the only solution.

Audrey: I'm a major fan of Piketty, but if you don't disrupt capital, it will ... We're just going to the place where undisrupted capital accumulation goes. And I do think thinking about things like expropriation, we need to seriously think about expropriating wealth that has been accumulated in highly unjust ways, and probably in some cases, ways that could be defined as criminal.

I think those more radical solutions, we need to find the space for that in this context, where the discourse allows us to put a more radical solution on the table. Because I agree with you, absent that, capital is looking for some way to make a return. Nobody's remembering 2008. When capital wants a return, they will pull something together until it falls apart, and then the public will pick up the pieces.

That's almost certainly gonna happen in some form again. We weren't prepared last time. This time we should be, and we should be prepared with more radical proposals, and I think those of people like Thomas Piketty. But I think some of the stuff Jason Hickel also talked about when he talked about degrowth as a movement, I agree, Lou, with what you were saying.

But there are ideas there. Are we ready to push them forward? Because there are going to be some moments of breaking here, not least of which is what's gonna happen in a couple of years' time in the US? It could get catastrophically worse. There could be moments of opportunity. We don't know. But I think sometimes we have our heads so down in the problem that we're in at the moment- Yeah

that we don't look at where are the moments going to come, and are we really ready with some of the things we've been talking about here?

Max: There is a moment now to, to get back at least some of the infrastructure, because one thing that Europe is known for and what- where we are very p- specific is that most of our infrastructure is state-owned, right?

Power grids are state-owned. Some of the utilities are state-owned. Our whole utility infrastructure for water, roads, 99%, I would say, is state-owned. And to this day, the fact that fiber is not state-owned, in Europe especially, seems to be a huge contradiction of our social contract and our model of how we operate in most European countries, and putting data centers into that same bucket and computation in that same bucket of, is this really a good idea to be in a private- privately organized market?

Um, I think that's where the sovereignty debate can get us towards not per se maybe expatriation, but recalibration of that market, maybe the state taking a certain share of it, or maybe using our pension funds to, to finance community-owned or re- rebuying of infrastructure, which maybe a few years ago was far-fetched, but I think that these more radical ideas are

Now is the moment for them.

Alix: Okay, so final thought. Maybe this is for a policymaker in Europe that has just said, "This data center project is really important for our national sovereignty." You're trapped in an elevator with this person. You've got 30 seconds to a minute to say why that is just bad. What would you say to this person?

Lou Welgryn: I would say the way we're developing data centers today, we're crushing our resources and exploding our climate objectives. We are not building any sovereignty the way we're doing it currently, and democratically, we are excluding all the citizens from those decisions. So the real question we should really ask ourselves is, who is making money from that- Private companies, extra foreign companies, we don't want that.

And what we should be asking ourselves is, what do we wanna do? And put social interest first.

Audrey: Honestly, if I spent 30 seconds in an elevator with a European policymaker, I would breathe deeply and think about how I could get information into the hands of a small grassroots group finding a data center somewhere, 'cause it would be a far better use of my time.

Max: I would probably say something like, "Assume you're getting all the cost and they're gonna be dishonest about giving you any value. Do you really know, and have you talked to your citizens that you are supposed to represent why you want this?" I talk a lot with regional policymakers about these things, and to be really honest, and I think it's something we for- sometimes forget also as NGOs, 90% are like, "I don't want it either, but the land is owned by a private individual selling it to a private individual.

I have no legal tool to stop this data center." Unless there is a law that gives me some kind of right to block a data center from a national law perspective or EU law perspective, they literally, they can set a policy around it that it should be green, it should be beautiful, but they, unless the, the government owns or the state owns the land, it's very difficult in our neoliberal way to block these private transactions.

And I think we should advocate for the tools that are missing for them to do this for us. Because again, I talk to a lot of them, and it's not that they don't want to, it's they literally don't, either don't know how or don't have the tool to do it.

Alix: A huge thank you to Audrey, Lu, and Max for joining us, and to Luminate for partnering with us to make these conversations possible. Next week, we're turning to Africa, where social media and AI are often framed as engines of economic opportunity. But those opportunities come with significant costs, from exploitative labor practices to environmental harm.

We'll be joined by Benga Sesan, Chenai Chair, and Ijeoma Ochai to explore how communities navigate these contradictions and what more equitable participation in the digital economy could look like. These episodes are part of a broader partnership with Luminate highlighting how activists, researchers, creatives, and organizers around the world are building new visions for how technology can better serve people and society over profit.

You can find out more about the series, including our upcoming live shows and streams, through the link in our show notes. Thanks to the podcast team, Kushal Dev, Marion Wellington, and Zoe Trout, and to Peter Price at Rethink Audio for all their work making this show happen every week, and we will see you next

time.

‍