Arun Kundnani (AK) interviews Seda Gürses (SG)- a critical computer scientist at the Delft University of Technology in the Netherlands. They discuss the ways that tech corporations such as Apple Google, Microsoft, and Amazon acquired new powers in public health as a result of the response to COVID, and the issues of corporate power and accountability that this raises. It’s an issue that’s generally been largely ignored, and Seda explains how these new developments do not fit earlier paradigms of privatization and algorithmic power. She brings some really thought-provoking ideas which raise profound questions about this new world being invisibly created around us. Seda works with a group called Programmable Infrastructures, which looks at developing alternatives to corporate computing. You can find her on Twitter with the handle @sedyst.
I have put in bold AK’s questions and what I thought were the more worrying aspects of Seda’s messages.
Transcription edited by Martin Blanchard April 2025
Power grabbing disguised as healthcare innovation
AK I just want to take us back to a moment in April 2020 when we were still just beginning to get our heads around COVID and there’s this moment when Apple and Google announce that they’re going to work together to put Bluetooth contact tracing into their cell phone operating systems. If you could just talk us through what that moment represented, what was the proposal there, and what its significance was?
SG Sure. I think it might be nice to first rewind a moment and say that Google and Apple were summoned by governments who said, ‘We should have apps to amplify our capabilities and contact tracing.’ And the reason they summoned Google and Apple, even though they didn’t state the names of these companies, is because there’s no way to deploy an app in the world today without going through the operating systems
of Android phones, which are basically controlled by Google, and Apple phones,
which have iOS which is controlled by Apple. So, in a sense they said, ‘For our public health activities, in this case contact tracing, we want to use the platforms of these two companies, therefore we are ready to sit at a table and negotiate with them.’
In a sense, one could say, ‘okay, but they could have just deployed an app.’ But indeed, this Bluetooth matter was relevant because basically Bluetooth was never meant to measure distance. Bluetooth was there to transfer files-it was a replacement for cables, basically a ‘wireless’ technology. What they wanted to do was to use Bluetooth to measure distance, and to do that they needed Bluetooth to work in the background and not consume too much phone battery power. For this they needed Apple and Google and at some level, although I don’t think many people recognized this at that moment, they needed the two companies to make their Bluetooth signals – that were going to be used for contact tracing – interoperable. They needed their collaboration. Therefore, the moment that the government said, ‘We want apps,’ what they really said was, ‘We want Google and Apple to be part of our public health initiatives.’
Apple especially, I think, saw this as a great opportunity to build on their growth mandate which included growing into other sectors including health. They had been making a lot of initiatives in the direction of health, and public health, which many people will know about because of all the ways that iWatch or the step-counters are already so symbolic of such initiatives. What also happened at that moment for the government was that their imagination of how they would get in touch with their population was hijacked (in an exciting way) by the development of mobile devices; these ‘things that we carry around all the time, that are in our pockets’ and that as a project, had been worked on since the 1990s. Government thought, ‘Oh my god, this is a great infrastructure, we can use it for public health, it’s like having Whatsapp in everybody’s pocket, we can have public health in everybody’s pocket.’ This is how Google and Apple entered the game: they did not ask to enter, they were called in by governments who wanted these apps. So that is point number one to remember.
Number two is when the governments first asked for apps, although a lot of consultancy firms and a lot of startups came in and said, ‘We’ll provide the app,’ government also wanted to collect all sorts of data that was absolutely unnecessary for contact tracing; they were proposing a lot of technologies that would basically build a massive surveillance infrastructure in the name of public health. As is known very well from 9/11, if these kinds of surveillance infrastructures get set up in the moment of an emergency they remain. Therefore, my colleagues and I, and many other communities who work on privacy said, ‘Okay, let’s do some damage control, and let’s see how we can lower the bar for entry [to provide a solution] for people who are proposing apps that collect the minimum amount of data to enable contact tracing’. Until this moment we thought that Google and Apple were probably going to come into the game, but we were functioning without them. Then they entered the game and said, ‘Actually we will take that privacy preserving protocol that you guys are working on, and we won’t just make it an app.’ In the case of Apple, they said ‘We’ll push it into our operating system, and we will silently update all people’s phones so that next to your sound settings or your telephone settings you would also have a contact tracing setting which is called the Google/Apple Exposure Notification. If governments want to build apps on top of that we will control which public health entities come in, and any apps will have to be government sanctioned so we’ll only allow one, or if it’s a regional public health or federation of public health institutions, we’ll allow them to have apps, but they will have to use our interface from the operating system to these apps.’ In doing so they started not only to decide where public health would end and where their platform power would begin, but they also pushed the grounds on who could do scientific work. For instance, at that time myself and my colleagues were working with epidemiologists who told us the distance and the amount of time considered to be a risk for somebody, so that they would be defined as a contact by the World Health Organization recommendations. This meant that there were massive levels of calculation that had to be done, and had Google and Apple said, ‘we will define those risk scores,’ they would also have put all that scientific work behind their infrastructure. What this tells you is that they suddenly had the power, sanctioned by governments, to enter everybody’s pockets and enable a public health function which would normally be a function that only governments would provide.
As a result, they were able to position themselves as arbiters between citizens and their governments. This was also very clear in the way that they said, ‘we will take this privacy preserving protocol, and if a government comes ―like the French government did― and says no, we want another protocol that would indeed collect more information into the hands of government entities or public health entities because they want to do research with the data etc,- we will say, ‘no you can’t, you can’t do that.’
By this, Google and Apple were saying ‘well, your government is kind of intrusive, and we will protect your privacy via-a-vis this government.’ And in this way they really positioned themselves both as inevitable or, let’s say, inescapable partners in public health for governments across the world that wanted to use apps and had the ability to use apps. They positioned themselves as arbiters between citizens and governments with respect to privacy. So those were huge steps, I think,
Big Tech’s Power increasing
AK Can you say a bit more about how that kind of positioning played out subsequently and did that lead to them to grabbing new kinds of power or new resources in terms of the public health space subsequently?
SG The interesting thing about big tech companies is that this is not just your typical privatization where you capture government tenders-which also happens. They practically did this [covid tracing] act for free. The question then is, what is their gain? I think the gain is in a sense twofold: they got governments to recognise that ‘if you want to use our infrastructure then you have to go through us, and our infrastructure is very attractive to you, because you can reach to, whatever, 60, 70, 80 or even 90 percent of your population depending on the phone penetration.’ And governments thought this was a ‘cool’ thing to do. A lot of politicians afterwards tried to look ‘cool’ by saying that they would make an app (for some random thing) and this then began to ‘normalize big tech infrastructure’ as a go-to for public services.
On another level big tech managed to ensure a seat at the table across countries for public health decisions. So now people would make sure that they ask Google and Apple what they thought or whether they would implement things etc. So the gain was not monetary, it was not a capture of resources per se, but it was a capture of infrastructure in the sense that instead of having call centres which would be funded by governments with people who were trained to do contact tracing, what we ended up having was an app that just did some calculation and which had quite a bit of randomness because Bluetooth was not made for this.
At best it led to call centres that you could get in touch with saying, ‘I got notified that I’ve been the contact of somebody whose test is positive, what do I do now?’ So that part you could see as more like classical privatization. But instead of having public health workers, you were all of a sudden forwarded to, in the case of the UK for example, a bunch of teenagers who had that morning received a message in their room in lockdown and using their own computers. They now became a call centre worker having to answer sometimes rather disturbing questions such as ‘I can’t breathe,’ ‘my husband is lying on the ground,’ or whatever- such things for which they simply were not prepared. In that sense it was NOT like a classical privatization. I think it is also very important to distinguish that the way big tech ‘moves in’ looks different because of the financial power they have gained over the last decade. Their power is not necessarily in the capture of very simple kinds of privatization, but in the capture of ‘the whole infrastructure’, which is just hard to imagine and quite, quite different.
How are large companies taking over the infrastructure of our healthcare systems?
AK Yes, absolutely, there is so much in there to think about, so let’s try and break that down a little step by step. Can you say a bit more about the kind of computational infrastructure that’s central to the story you’re trying to tell? What do we need to understand and how has the COVID response been built on that infrastructure?
SG I mean something very, very specific with computational infrastructure and,
in the fields that I touch on as a computer scientist like information science and studies, people have been talking a lot about data infrastructures building on the work of Geoff Bowker[1] and Susan Leigh Star.[2] But what I really mean is the material infrastructure and specifically, I’m referring to the ‘clouds’.
So, for many people who know a little bit about technology, and they have used either something like iCloud or they have used maybe some Amazon based services, or Microsoft Office which is now Office 365, or Google Docs, these are all programs that are running on the ‘cloud’.
The biggest ‘cloud’ providers are: Amazon (AWS), much greater than any other company
by far, followed by Microsoft which used to be a software company, but practically in the 2010s they pivoted to become a ‘cloud’ company. The reason they did this may also be very relevant.
If you were around in the 1990s the business of computing was still organized very much around personal computers or on-premises servers, and software was something you bought in a box. Instead of queuing up to get iPhones people queued up to get Windows operating systems, and you would get a box. Depending on how old you are, the box contained either diskettes, CDs or DVDs and then you would install it, and sometimes it would work and sometimes it would not. With the 1990s and the rise of the internet, competing companies to Microsoft really used this advantage to deliver software through the net. This affected the production of software in that instead of the massive distribution of boxes, which meant you had to write the software and then say, ‘okay, today we’re done, we’re not going to do any more corrections or changes, we’re going to slip those discs into a box and send it out,’ they could now deliver software over a browser. You could click a button, and the software would load, and it would be on your computer. Now, there was still difficulty in maintaining the software so that was also not so great. Therefore, little by little since the 1990s what we have seen is things called Web Services or Service Oriented Architectures where software is no longer brought to you discretely, but instead you use the software that is on somebody else’s server. Like Google Docs vs. having Microsoft Word on your computer-this is a huge difference. Aside from the fact that your files are now on somebody else’s server-you might not be able to find them if you’re not online. It also means that every action you take with the software is trackable, and that tracking has allowed software providers, which are now service providers, to track user response and to continuously optimize the design so that they can get the response they would like from these users.(This is not like the kind of manipulation that Shoshana Zuboff writes about, but it is making an operation out of delivering software and services that are optimized at the bottom-line of the company using the clicks as a sort of feedback into the operations of that company). ‘Clouds’ therefore became the way services get delivered, because if you deliver software through services and you have to have software running on a server, you now have a cost, and it is that cost that got turned into a business model and became the ‘cloud’ – to provide the use of servers and use of ‘on cloud’ software, at a cost to large numbers of other companies so that they may start providing their services to millions and maybe billions of people [MB and ‘cloud’ owners maybe also to take a percentage of profits].
I am telling you all these details because they eventually become important for what kind of imagination ‘cloud’ has of a society. For ‘clouds’ you can think Amazon, Microsoft and Google which is trying hard to catch up. Of course, there are other smaller ‘cloud’ companies which are also relevant, but in terms of the financial boost that these companies got A, M and G are the main three.
Then with mobile devices, which we see as accessories of the ‘cloud’, and I say that because we have moved away from the personal computer that you manage and maintain yourself, to holding devices that are basically maintained by the companies. This is why Apple could do a silent update on your phone and push contact tracing onto it. In a sense it’s not a personal device, it’s a managed device that is an accessory on the ‘cloud’, and most of its functionality is on the ‘cloud’. You use your phone, and if you go through a tunnel your internet is gone and none of your apps work because all of them are running on the ‘cloud’ somewhere.
For me, computational infrastructure is the ‘cloud’ plus these mobile devices as their accessories, and they have become in a sense the default place for software production. What I mean by that is, just like governments could not have apps without going through Google and Apple, if you’re in a competitive business it’s practically/ economically impossible not to go through the ‘cloud’ plus mobile model to deliver software to people today.
The social consequences of ‘cloud’ computing
AK. So, they have captured software production globally, and what does that mean in terms of the social consequences of ‘cloud’ computing of power relationships? What does this mean in terms of the social and political consequences for those of us who are walking around with these devices in our pockets and are kind of plugged into this infrastructure?
SG I would love to know the answer to that question, but I think that IS the question, right? We can try to touch on it. First of all, what you have is what could be considered a massive infrastructural power in the sense that if you are an organization that wants to have contact or have access to people, either as a government [or a public service] or as a company that wants to do advertisement or news agency, then the most powerful, and potentially the default infrastructure, is to go through this ‘cloud’ and mobile space that I have just mapped out, and which is in the hands of a few companies. I think that itself is already massive power.
The fact is that this infrastructure is of course organized in a particular way and optimized in a particular way, and is subject to financial pressure, which means that the infrastructure power ‘in and of itself’ becomes something we must understand and keep an eye on, especially because it is usually invisible. We do consider data and algorithm issues, we do consider the media, but we must also consider the way an infrastructure is built (in this case it is this computational infrastructure) because that is what determines what it means to access it, and the conditions under which that access is possible as determined by the four or five companies.
But they themselves are now also determined by their infrastructure. The design is ‘sedimented’ and I think that’s really important.
The other element is what we call ‘pocket power’ which is that even if these companies cannot get organizations or governments to accept certain services, they can enter them through our pockets.
We saw this at my university, where our IT people were smart enough to say, ‘look, Zoom, for privacy and security reasons really is not a very appropriate software to be used in an educational setting, and we should not use it’. But then the professors came and said, ‘But I use Zoom, so could you guys get us some licenses anyway?’ If you think about it from the perspective of the business of computing, enterprise software[3] is not trivial -which Microsoft has mastered and to some extent, so has IBM. It is in a sense, much easier if you have a service that is already accepted and used by millions and billions of people and then those people bring that service, and make it normal in an organization. That has happened with Gmail, that has happened with Google Drive, that has happened with colleagues communicating over Whatsapp, or indeed professors asking for Zoom. I think we should not underestimate this ‘pocket power’ which is basically using what the industry has focused on for decades; the idea of ‘the user as king.’ You must have a great experience and all other stakeholders do not matter. They have really managed to push this, through things like Agile Software Development Methodologies, as a way to enter organizations which, to achieve otherwise, they would have to go through the usual tender processes as part of the contract process, and etc. So, there are different ways in which this power shows itself and I think it really is an interesting question to study how that works out.
Governments and big tech regulation
AK Can you say a bit more about how this looks then in terms of the relationships between governments and these corporations like Amazon, Apple and Google and so forth? If I’ve understood you correctly, you’re saying that the kind of infrastructural power that these corporations are at the helm of has given them a seat at the table in things that seem at first sight to be unrelated to their core business, like for instance public health. Can you say a bit more about what that relationship looks like between those corporations and governments around things like public health? You said that we can’t understand it in any kind of conventional sense of privatization. Can you try to paint a picture for us of how you would describe that a little more?
SG I think it’s not a simple story, and I think a lot of my colleagues see this as that ‘there is this corporate power coming through big tech, and the governments are separate and try to regulate them, or act as maybe a counter power, etc.’ However,I don’t know if this is seen that way in other fields, or by other activists or political circles. But I think what we saw with COVID was different.
We did some measurement studies to look at educational institutions-in which we saw what we could call ‘the cloudification of public institutions,’ as an ongoing process. While this is not just a COVID project, I think the Pandemic maybe made it happen faster. Indeed, if you look at the bottom lines of these companies it definitely made it faster! But it also maybe, exposed some of the power of these companies: they were taking over institutions at a massively fast scale.
So, there are a couple of things that we need to think about.
One is that governments are understanding this power [that big tech brings] and are kind of excited about it. When we were building the contact tracing apps, for example, I would get calls… I got calls I think a couple of times…where government entities would ask, ‘Could you limit the number of notifications people get? Could we mess with that?’ Because we had a very concrete definition, if you’re a contact you get a notification. But they said, ‘If we don’t have enough beds maybe we don’t notify people because we don’t want them to go to the hospital.’ But if you look at the fact that some of these governments that were calling have invested no money into public health since COVID, and that sometimes they have even cut down the number of beds, you can see how the possibility to optimize the number of beds versus notifications as a form of public health is just very yummy for them and very scary for us. I think we saw in a sense, governments being excited about using this infrastructure for their own interests. We saw, I think, these big tech companies saying, ‘Let’s work together.’
The government excitement derived from that ability to very precisely manage populations and reactions and behaviour through this infrastructure. I mean it’s a form of population management. I must give a ‘hat tip’ to my collaborator Martha Poon with whom I discuss a lot of these things and she’s not here now. Most of the things I say can be attributed to her, but I think that these infrastructures enable a certain kind of operational control, and if they do so in a, well…if they do well in that, which means that it reduces cost and it has some sort of, like…..it fulfils some requirement for effectivity, then they [big tech] can capture an existing infrastructure.
As an example of this having already happened, you can look at Facebook. Although they don’t really have as much of power as the other four companies like Amazon and Microsoft with the ‘cloud’ and Google and Apple with their mobile devices, Facebook has become the go-to for advertisement and for accessing people.
What they have done is to continuously ‘super refine’ their operation for advertisement to become the default place for all advertisements. They have captured what was maybe not even seen as an infrastructure but that did exist, which used to involve ad companies, and maybe the press, different publishing houses, etc., or TV broadcasts, or other institutions. Now all those parties go through Facebook, and they not only have to give a cut of their income to Facebook but they are completely dependent on them to act in this field as publishers or broadcasters or news agencies, etc.
So… I think if for a moment we pause and say, ‘The power from this whole thing does not come from mass profiling individuals, but it’s about refining an infrastructure and making it the default, and shifting whole industries or sectors onto it for that operational need. Like for delivering ads, in the case of Facebook and Google, for delivering goods in the case of Amazon as an e-commerce website. But also, for delivering all software in the case of ‘cloud’ and the mobile devices as accessories.’
You then start to understand the shape of this thing and understand that it’s not about individuals as much anymore. Instead, it is creating an operation through which populations can be managed. The emphasis is on operational capture, and maybe we can say that the capture of the people, while it can be refined, is kind of done.
Tech companies in Kenya and the Global south
AK You mentioned the UK and France, tech companies in Kenya and the Global South. Can you say a bit more about how this might look in different parts of the world, or is the story you’re telling something that’s a European story? How does it look elsewhere?
SG I think that’s also another interesting research question. In collaboration with the Institute for Technology and the Public Interest–which includes myself, and also Miriam Aouragh, Helen Pritchard and Femke Snelting– we are asking some of these questions such as ‘how are things different in different situations?’ We are currently in the process of asking ‘what happens if these government entities are not there or have been failing or struggling?’ Specifically, we have been asking ‘what happens, for example, in Lebanon after the multiple crises that they have been going through over the last couple of years?’ Or in another project we are asking ‘could we bring contact tracing apps to a couple of countries in Sub-Saharan Africa?’ There we see that not only is the position of these tech companies different but also the position of the government and civil society, and knowledge about these companies is very, very different. So, we are asking, ‘what does it mean to capture operation when the state is kind of failing, and then you really become the default?’ Also, ‘how much do these companies need these governments to uphold a certain position so that they can capture operations?’
In the case of Kenya, what you see is that a lot of tech companies have chosen Kenya as an experiment ground, which means that they can do highly unethical and politically concerning stuff and get away with it in a way that they would not in Europe. Then the question is, ‘would they get away with it in the US?’
So, I think, yes, those things are very different. The other thing of course that is very different is that the ‘cloud’ companies I have so far mentioned, do not account for all the companies that are very powerful. We have a bunch of them also in China, and, for example with the war, I have seen some information from Russian computer scientists, I don’t know what school they are, but they say that, ‘Russia has already been looking at ways to exit the infrastructure that is in the hands of the four US companies, and here’s a plan. And we are going to collaborate with China to do that.’ So I think it’s going to be very interesting to see how these regional manifestations differ. But of course, these big tech companies like to project themselves as universal, global, and somewhat comparable wherever they go- which is why they can say, ‘we’ll do contact tracing across the globe.’
The future of health and technology
AK And so, can you say a bit about the future of health and technology; how does this look going forward? You’ve talked about how these tech companies are not necessarily thinking about immediate profitability on some of these actions, they’re building up a kind of positioning, relationships with governments, capturing a certain space. But where do they go with that over the coming years? And specifically, around public health, what does this look like in terms of how we might see public health evolving, say for example in Europe?
Europe’s an interesting place for this question because probably more so than anywhere in the world it has a certain… it’s not always upheld, obviously, but it has a certain sense of ‘norms’ of public health, and what that means in terms of a democracy and so forth. Rights that people have fought for and established especially over the course of the 20th century-the NHS in Britain and so forth. This is certainly more so than in the United States and most other places in the world. Is there a way that these developments that you’re describing may impact on that?
SG Yes I think so. There are multiple things that could be said. First, I think the ‘cloud’ has an ‘imagination’ of organizations. The idea is that an organization is no longer an institution that gets IT support, but instead an organization ‘sits on top of the ‘cloud’ and bundles a bunch of services, and stretches it out to their constituents-in a sense, this is so normalized already. We taught a class, and we asked students to study six companies and they chose them depending on what they were interested in. It was just fantastic to see that Uber basically exists on the ‘cloud’. Or a company like Netflix practically exists on top of Amazon, even though Amazon has Prime. It’s amazing. But I think what big tech would like is to have universities become little shells that exist on the ‘cloud’. It’s just a bunch of digital services like whatever MS Teams plus Zoom plus whatever HR service for hiring, etc. Then you can basically gut out much of the administration and much of the operational makeup of organizations and move it to the ‘cloud’ and use the organization just to keep on stretching the ‘cloud’-in the case of the university to students, and in the case of the health system to patients, and public health system to the people that public health is supposed to serve, although they don’t always.
I think the imagination is really, just to keep on doing that. If we think about Facebook and Google as having made their billions, if not trillions by capturing everybody’s advertisement budget, this thing is up for capturing everybody’s IT budget. But Information Tech is supposed to be everything that an organization does, that is except for that little shell that says, ‘we’re the university, and here’s our flag, here’s our president, and here are the professors who are also, of course, using services too…’
I think that’s the imagination, at least organization-wise. Also, I think there are
other imaginations that I don’t necessarily study but which are about managing disenfranchised populations. Basically, instead of making money from individual users you get a government to pay you when you manage, let’s say, a refugee camp, because you have the phones, you have the whole infrastructure, or you give them cards or whatever, you give them some dongles, and then you can manage.
Everything that’s logistical that can be digitized they can manage. Then all the expensive and complicated stuff will be pushed on to governments, or whoever else, to take care of, and if nobody’s taking care of it that means ‘we’ (big tech) will have to take care of it. So, I think it’s interesting to see how they capture, and indeed it’s not just like a straightforward privatization but is instead a complete reimagination of the organization as a structure. I think maybe that’s a good starting point.
Then what did we see, when COVID started with the lockdowns especially? (I don’t know how this is going to play out.) We saw how it split the society first into two, and then maybe three, and possibly more parts. But generally, it created a delivery class and a receiver class. The delivery class was those care workers who were delivering the care-doctors and nurses and what have you-and those who were delivering packages-all the people who brought the Amazon orders and whatever to the home. The receiver class were receiving orders either to keep business going as usual but from home or receiving their personal packages from these online services. If you can imagine that in an ideal world, Amazon would not need the shops, Amazon just needs to deliver. I think that was a kind of quick exposure into the kind of society that they imagine with of course, certain populations falling out of both deliverer and receiver roles, and they are ‘the policed’, and ‘the managed population’. That is where we get the refugee camps or people without shelter, who are then absolutely surveilled and policed using big tech technologies.
It is kind of interesting to think about how the ‘cloud’ imagines us so that we can develop counter imaginaries; I think these were like little windows into the imagination of a society [as seen by] these ‘cloud’ companies with mobile devices as their accessories.
The dangers of technology
AK Absolutely, and so, the story that you’re telling it seems like it’s something that hasn’t been focused on enough at all. Even just thinking about the COVID question. Do you think the left and liberals have failed to really see some of this stuff? Because we’ve kind of had this desire almost for the state and for tech corporations to be the solution in this moment? We want them to save us as it were. So, we have not been as questioning, perhaps as we might have been, of some of the political questions that are implied in what you’ve just been describing.
SG I think that there are many reasons why we landed here. I think there was a lot of excitement about the internet, it was going to connect us all, and it was going to be this amazing thing. There were always voices like Paula Chakravartty and Miriyam Aouragh saying that this is an imperialist infrastructure. But we had then people at Berkman saying, ‘It’s gonna free us.’ So, who to blame? We were all excited about the internet, but I think what happened, and this is again Martha’s wisdom, is that the business of computing kind of crashed into the internet, then kind of crashed into the global financial system that was looking for good investment objects after 2008. There was a strong relationship between finance and computing from the beginning, and I think they built something on top of the internet that has become invisible to us.
I think that if I look at some of the critical work in the field that I touch on, or that has now also bled into computer science, a lot of the consideration has been about data and data collection and surveillance, and not about capture, infrastructure build, and what it means for political economy. So that work was not done, and instead we were very worried about privacy, including myself. I worked on privacy for many years, and then most recently on algorithms and automation and what it would do to labour, or algorithmic discrimination. All those ‘framings’, which I have to admit were not only coming out of critical scholars but were very much supported by industry in the research centres that were continuously budding, and that were originally getting half a million or a million or two dollars from Microsoft or Google or whatever. These were really pushed and became [still are MB] the main frameworks through which to look at technology.
Technology was going to be innovating, and it had edges, and those edges could be captured with data and algorithms, and we could either tweak the algorithms or shave off some of the data and then it would all be fine. But that was without looking at the fact that these were massive infrastructures with their own imagination of what future society will be, and with an infrastructure that is so strong that it even drives the engineers. You cannot as an engineer just go and change this infrastructure, it’s already established in many ways. It’s just like saying we’ll take cars off the roads; like what are you going to do with the roads? I mean, I’m sure people have some imagination of what to do, but what I’m trying to say is that the infrastructure has been built for decades and decades and it’s so settled and so invisible to us.
I think not seeing it has been to our detriment, and I really hope that more people ask questions about it, and look into things beyond how people are having agency vis-a-vis algorithms. That they start asking ‘What does it mean if the university (or any public administration, such as public health or healthcare etc.) is just a little shell sitting on top of the ‘cloud’?
AK Are there ways that we can imagine making demands of say, local government, municipalities, or public health departments to adopt policies that might at least contain some of this, or limit some of this? Is that something that we can imagine?
SG I think, to connect the previous question to the one now, a lot of governments have gone for a ‘cloud-first’ strategy (laughs), meaning that they are literally pushing ‘cloud’ in their countries because they have been told by consultants that this is the way to go with technical innovation.
COVID really activated governments across Europe and probably to some extent in the US-although as you said public health looks different in the US-to digitize the public sector. They have really just been caught by the consultants that are pushing ‘cloud, cloud, cloud’, all the time. I think it’s going to be a big battle. It’s going to be a very difficult thing to push back against because we have a multi-trillion dollar industry saying get on the ‘cloud’ under terms like ‘digital transformation’.
It’s another podcast probably, but ‘machine learning’ is for me absolutely a driver to the ‘cloud’. We did not make massive breakthroughs in computer science, no, we just simply managed to make computing very expensive. We found a way to do computing in a very compute heavy way, which is ‘machine learning’. And, the best place to do it is?… surprise! it is on the ‘cloud’. So, I think the push back is going to be very hard.
The push is there, but it hasn’t quite happened yet. I think there are at least two levels that are really going to be important. One is to be very keen on hearing about ‘digital transformation’ in our organizations. This could be public institutions or any organization actually, to ask what is the vision of this institution or organization that this ‘cloud’ or ‘digital transformation’ is going to serve? Because right now we are doing the ‘digital transformation’ for ‘digital transformation’ because it’s going to ‘innovate’ the organization. What this actually means is that the organization is open to change. But we need to turn the question around and ask ‘okay, if you’re open to change let’s talk about what that change is, and make sure that ‘digital transformation’ serves the organization’s vision instead of the organization serving the ‘clouds’. I think that’s a very important direction to go.
Then, with respect to everyday practices I think it’s really important to talk with institutions (and this is a really difficult one, since health institutions across Europe have been really struggling with austerity, that they have been subject to over the last few decades) to ask ‘what is your practice and how are you going to maintain it, or, what does it mean to change your practice so that you can serve people better? And how then do you ask for technology that allows you to do that?
I can give you the example of immunity passports. I think they are basically these certificates that you get if you have had multiple vaccinations, or if you have had a test, or if you have recovered from COVID-they of course also went into the mobile phones we discussed earlier. When we talked with people in public health, Martha and I explained it to them as follows: we said: ‘your job in public health is to inform and enable people to make good health decisions. When your practice gets translated into the phone to certificates it becomes a practice of checks asking, ‘have you done this?’ That is a complete transformation of your practice. So, let’s talk about this. Are these checks what you want your practice to be or is it something else? What does it mean for somebody standing in an airport, somebody standing at the door of a restaurant to be checking these things with no public health knowledge? What does it mean to have your workforce replaced by a bunch of people holding phones and doing checks?’
People need to be very diligent in the changes that happen to practice because, as I said, this infrastructure has a very specific understanding of how we interact- of course you can expand it, and you can design systems differently, etc.- but it has a path that it has already selected, and if it doesn’t serve you then why are you going on the ‘cloud’?
These are very simple questions that do not get asked. If you just look at data and algorithms, and if you just look at algorithmic discrimination or what have you, well these are all actually ‘frames’ pushed by industry to avoid regulation and to make it seem like by tweaking algorithms they can solve all our problems.
AK Thank you Seda. That was really helpful. There’s so much stuff here to think about and you explained it really clearly. Thank you very much. Great.
SG Thanks. Thanks for having me.
[1] https://estsjournal.org/index.php/ests/article/view/1253/629
[2] Bowker, Geoffrey, and Susan Leigh Star. 1999, Sorting Things Out: Classification and Its Consequences. Cambridge, MA: MIT Press.
[3] software designed to meet the needs of an organization, rather than individual users, that is crucial for managing business processes, operations, and data.
