Sigrid Keydana | Dissecting the quick fix: Analysing tech-solutionist solutions | RStudio (2022)

Hello and welcome to my talk, Dissecting the quick fix, analysing tech-solutionist solutions.

Let me start with a definition. What is tech-solutionism? Tech-solutionism is the tendency to use technology to solve a societal, political, organisational, environmental, what have you, problem. A few examples should make this more clear.

Three examples of tech-solutionism

Example number one, riots. What's the problem? Well, suppressed population groups, for example, ethnic minorities, are starting to mobilise. What do we observe? This is more likely to happen when it's hot, very hot. What is the solution? Well, just install climatisation throughout. And what do we gain? We don't have to actually address those groups' demands.

Example two, mass shootings. What is the problem? Since 2014, on average, there has been more than one mass shooting a day in the US. What do we observe? You can't always see if someone is carrying a gun. What's the solution? Well, just install advanced surveillance technology. And what do we gain? We don't have to change the laws related to gun ownership.

Example number three, climate change. What is the problem? Climate change is rendering parts of the earth inhabitable and strongly impacts living conditions and others. What's our observation? Time is passing and we are less and less on track to keep the situation under control. What's the solution? Well, we just move to Mars or somewhere else in space. What do we gain? Well, we survive and get to have fun, if that's the kind of thing we find fun, right?

Why cynicism isn't enough

So yeah, I think by now it should be pretty clear what I'm talking about here. On the other hand, you probably also have noticed I got more and more into some kind of cynical tone, like disgust, cynicism, and pure cynicism and just talking that way doesn't bring us very far. Why?

Well, on the one hand, we won't notice less latent solutions, quote, unquote, when it's less clear. Like, for example, talking about climate change, what about things like geoengineering? Isn't that easy? And also, we're not never going to reach those people who don't already agree with us anyway, right? So I don't think this is very useful.

On the other hand, I think what we should try is to understand what is so seductive about this thing. Well, like the word already says, it solves something. Good, right? It solves a problem. Often it solves a hard problem, a problem that otherwise we wouldn't know how to solve, how to address even. It solves a problem we would have to make sacrifices for. And we don't want that, right? And often it just solves a problem we're terribly scared about.

A framework of W questions

So what I'm going to try to, in this talk is, or what I think we need is some kind of framework to analyze those proposals, and a framework which does not just ask the technical questions, because this is something we are already doing to some degree. But I think the technical aspect is not enough, and so I'm going to try to propose a very simple framework, a framework of W questions, like what, where, when, why, who, to whom, how. It's not a W anyway, but to address this in a systematic way. And there I want to focus on the more neglected aspects.

But as I'm going to focus on two of these aspects, let me say what I mean by the rest of them. First what. What does the solution do? And this should include the complete technical workflow, like from where do we get the data from, how do we train our machine learning models, how do we evaluate them. And in this area, we already have some things like data set documentation, model documentation, which is all great, but which is all the technical side of things.

Then, where, when, and to whom will the solution be applied? And this should include the complete geographic, social, political, et cetera context. Then, how will the solution be applied? And here, really, the details matter. Next, who is implementing the solution? There, as you see, I want to talk about more. And also, why is the solution being developed? There also, I would like to go into some detail.

So in other words, in this talk, I want to go into the why and the who, because in my view, these are the two aspects which are being neglected most in this area.

The why: motivations behind solutions

First, let's start with the why. Why. So what kind of things you're going to hear? Well, we care, we want to help, we are here for you. What kind of things won't you hear in this area? And here I'm kind of splitting it up a bit.

What kind of reasons might individual people have when they implement something or suggest something which they won't say? Well, for example, if I build this thing, it's going to look on my resume. If I build this thing, it might just get me a job at company XYZ. If I build this thing, I might be able to create a startup. If I build this thing, I'm going to learn that crazy new technology, X. And also, it's just so cool that this can even be done.

Next, in a more indirect way, when organizations do something, propose something, what are the things they don't say? Well, if we build this, we are going to strengthen our existing market monopoly. If we build this, we're going to distract from that bad thing that recently ran through the news. If we build this, we're going to be the good guys.

And now, in a more indirect way, let's think about side effects. If we build this, we are going to be able to extract valuable data for our targeted advertising. If we build this, we are going to get data to feed back into our neural networks. If we build this, we are going to have the foot in the door. If we build this, public institutions will be stuck with us.

Now, let's step up one level and think this is about technology. And technology is something that shapes us, that forms us, our daily lives. If we build this, we define how people live. If we build this, we define how people think. If we build this, we define what people feel. If we build this, we define who people are.

If we build this, we define how people live. If we build this, we define how people think. If we build this, we define what people feel. If we build this, we define who people are.

The who: a public–private dimension

By now, I think it should be clear what I mean by motivations by this thing, the why things are built. Now, let's switch to the next thing, the who, who builds something. And here, I want to be a bit more clear from the outset who I'm talking about because this can be a bit ambiguous. Who could also mean, like, to whom the solution is being applied? I'm not talking about that aspect right now. What I mean is, spelling it out, who designs, implements, and hosts the solution, who processes its data, who knows how it works, and can fix problems.

So let's take an example. As an example, I'm going to talk about education technology, also abbreviated as edtech, which has been booming a lot over the last years, especially, of course, due to the pandemic. And okay, with edtech, what kind of things are you going to hear? Well, with adaptive technology XYZ, individual learners' needs are being met. With adaptive technology XYZ, teachers can focus on doing what they do best. With technology XYZ, we make learning more effective and efficient.

What kind of things won't you hear, how it works? With technology XYZ, we are going to extract, store, and process lots of very private and personal data. With technology XYZ, we are constantly monitoring your behavior. With technology XYZ, you learn, we are learning from you.

Now this is about education technology. What's education? Education is like forming people, shaping young people. Yeah, forming people, let's just say it like this. So what this really means is we define how people learn. We define what is learned. We define what deserves to be learned. We define what deserves to be forgotten.

And stepping up a level, somehow a bit more abstract, what is learned? Common knowledge. This is building a culture. This is like who we are. So what does this mean? We define the past. We define the present. We define the future. We define what it means to be human.

By now, you may be wondering, can you say again how this is about the who? Yeah, and so before I was saying what I mean by who is who designs, implements, and hosts the solution, who processes and stores the data, and who knows how it works.

This who, or let's say like these who's, we can put them on a dimension. And here, I want to ask you to go with me and imagine a dimension, a continuous dimension with on the one hand, we have the public, and on the other hand, we have the private. On the one hand, for the public institution, I'm asking you to imagine an ideal democracy. We all know there is no ideal democracy, but we can all strive for one, at least hope for one, perhaps.

So on the one, this ideal democracy, on the other hand, for the private institutions, this can be anything. It can be a corporation, a foundation, an NGO. It doesn't really matter the official status. What are defining features of these who's on that dimension? For the ideal democracy, we have a democratic decision process with participation ranging on top. For the private end of the dimension, we can only say what's lacking, like this democratic decision process. Otherwise, it could be anything.

What are defining values? For the public institution, these are things like fairness, justice, human rights, accountability. For the private institution, again, it will depend on that institution's agenda. So now, if I assume this dimension of who, what kind of difference would that make? Well, for one, public institutions don't have to gather data to make money. Public institutions don't have to keep their trade secrets. Public institutions don't need to worry about market position.

Next, what difference does this make in political and legal respects? Public institutions can be held to ensure transparency. Public institutions can be held accountable. Public institutions can be obliged to be fair.

And again, stepping up a bit in abstractness, what difference does it make to human rights and values? Public institutions can be held to prioritize values over feasibility. Public institutions can be obliged to respect human rights. Public institutions can be held to promote, not erase, individuality.

Public institutions can be held to prioritize values over feasibility. Public institutions can be obliged to respect human rights. Public institutions can be held to promote, not erase, individuality.

What can we do?

Now, you may be wondering, still, I'm not a CEO. What can I do? Well, I don't really know. But I think one thing is perhaps we can make choices. This will depend. Perhaps we can make choices like who we work for, what we vote for, whether we organize politically. Or one step back, perhaps even, we can question the things we are being told, pretty much in the spirit of this talk, asking questions. Or even perhaps only we can start with a bit of imagination.

Maybe data science could be different. What if data were used not to perpetuate but change the status quo? What if data were used to make social relations more egalitarian, more just? What if data were used to subvert privileges, not reinforce them? Thank you very much for your attention.

Maybe data science could be different. What if data were used not to perpetuate but change the status quo? What if data were used to make social relations more egalitarian, more just? What if data were used to subvert privileges, not reinforce them?