Qué es la ética de la tecnología y por qué debería importarte | Gemma Galdón-Clavell | TEDxZaragoza

Good morning. I came to talk to you about the ethics of technology, what it is and why I humbly believe it should matter to you. It's a good week to do that because surely you all remember, you still have in your brain the images of Mark Zuckerberg, the CEO of Facebook, speaking before the US Congress, justifying the unjustifiable, which is basically

Facebook made us believe that we had some levels of privacy on its platform, which we actually don't have. But not only that. Luckily, today we are less naive.

in relation to technology. And we already know that when we don't pay for a technological product with money, in reality, we are the product. And our data is what is feeding a machinery, many times, of ads, a business model based on the commercialization of personal data. Today we also know, we are less naive,

that everything that is smart, all the smart technological tools, what they do is spy on us. And if you don't, I challenge you, if you have a smart TV, to read the instruction manual, where you will see that it is recommended not to have private conversations in front of your TV. If you have a TV in your room, you still have a problem. Because basically, a smart TV, like everything that is smart, spies on us. The product is us.

And in technology, I think that, fortunately, we have already gotten used to hearing people talking about the privacy issue, about data. But the ethics of technology go much further than privacy. And allow me to take a step back to tell you, to share with you how I started to realize that in technology we had to look at a whole universe of things that very few people were looking at. I started doing my thesis many years ago.

My doctoral thesis was on video surveillance. I wanted to see how the cameras impacted on how we used the city. I wanted to see what the motivations of the different actors were to trust a technological solution to a very important social problem, such as the perception of insecurity.

And there I was doing my thesis, discovering very interesting things, looking at metrics, seeing how cameras were able to capture social dynamics and expectations, but also to give them back to society, many times with differences. So the camera was not just a technological actor, it was not just a gadget. The camera intervened in how we used public space, the expectations we had about each other, how we understood security. But in all this process,

One day I thought of taking a book by Richard Sennett, a well-known sociologist, "The decline of the public man." And in it, he asked himself, long before we had drones and sensors everywhere, and cameras everywhere, and constant eyes watching us in our urban outskirts, he asked himself, "In these cities that we are creating, full of eyes, where would Romeo and Juliet fall in love?"

And that was my aha moment, my moment of realizing that while I focused on social dynamics and metrics and expectations, I had forgotten a very important question: Is what we are doing desirable? Do we want our cities, Romeo and Juliet, remember, children of two families facing each other, to have a forbidden love?

Do we want to create cities where young people from different families or different backgrounds cannot fall in love? And above all, thinking about young people. One of the things that has surprised me the most in my work is the total ignorance we have, ignorance to put it mildly. How do we ignore the data and the rights of young people? Young people appear in my work systematically as a collective

constantly mistreated by data collection. Young people, because they don't have their own home, they have a room, in the best case, where they put up padlocks and posters to avoid entry, a private area, etc. Their private area is often the public space, and we were taking it away from them. We were taking away from young people the spaces to give the first kiss in clandestinity, in privacy, in intimacy.

that they need and deserve, but not only for young people. Senec asked where Romeo and Juliet would fall in love, and the big question for me is: all this technology we're doing, all this money we're investing, is what we're doing around us desirable? Things that happen with technology, for example, linked to young people and urban environments,

For the first time in history, and as a consequence of the penetration of technological devices, it is more expensive to forget than to remember. Throughout history, it was much more difficult to remember. You had to write, use parchment paper, then print it. That has a cost. Right now, the difficult thing is to be able to forget, because everything generates a digital record. I always say that I am lucky to be over 40 years old.

because the first 20 years of my life, I don't take pictures of anyone. I have a picture, but nothing more. We condemn young people now to a life of fear for how their digital footprint can impact their lives in the future. That picture of that party, that comment on a social network, that moment we want to forget, that conflict,

even that judicial sentence or that file of a cause. All of that, people who have been born in the current society, totally digitalized, can emerge at any time and ruin your life, ruin your relationship, ruin the outcome of that work you so much wanted. Therefore, in my work on video surveillance, I realized that there were a lot of questions about technology that we weren't doing and that it was very important

that someone else would do. Since then, I've been lucky enough to dedicate myself to asking these questions, to explore what questions we should ask and how technology should respond to these questions and understand and incorporate all these concerns. I wanted to briefly detail some of the projects I've been involved in so you can see how this issue of technology ethics lands. I've worked for a long time, for example, on the automation of border crossing.

I don't know if you've been to an airport recently, but you'll see that many times the person who controls your passport is not a policeman, but a machine. You put your passport in, they do a digital fingerprint recognition, a biometric match with your ID, and the machine decides if you can or cannot pass.

When we started working on smart borders, one of the first things that surprised us, and that we later saw reproduced in many other cases, is that when the European Parliament started to propose the automatisation of the border crossing, the debate did not pass through the European Parliament because it was considered a technical amendment.

Changing police officers for a data process was nothing more than a technical amendment. It did not affect, according to them, according to the people who initially proposed that change, it did not affect the conceptualization of what the border was, so it could be a technical amendment that did not require any kind of public or political control.

To show you how little technical this is, the incorporation of the automation of the border crossing has made the border disappear. Nowadays, crossing the border, the line on the ground, the airport, is not important. We start to travel and we are allowed to travel when we buy our ticket or when we request a visa to travel.

and our step to the border doesn't end when we cross the border or when we reach our destination.

If we have asked for a visa, it continues because the validity of our visa is controlled until we leave the country, or if we don't leave it, an alarm goes off so that the police can take action. So the introduction of data, the automation of this whole process, has completely changed what a border means. The border is now a very large space for many people, a space of continuous control. Other things have happened.

With border automation, we have incorporated biometrics as a tool to verify identity. What happened in the past when someone wanted their identity not to be known? They burned their passport. When someone didn't want to be returned to their country because they thought that in that country they could suffer undesirable consequences, they could burn their passport and with that aspire to stay in a legal and legal space. Today, our passport

are our fingers. What are migrants doing? Burning their fingers, mutilating. Because it's the only way to prevent their body from becoming their enemy. What they wanted to happen as a technical amendment is having brutal consequences on the fundamental rights of migrants. Initially, or not, someone wanted us not to talk about it, not to ask ourselves what the implications of that decision would be.

Not only that, through my work I've been able to talk to more than 1,500 people who have used these automatic identity control systems across borders. And it's curious to see how when you ask third-party travelers, not Europeans, not people from the Schengen environment, what do you prefer: human police control or automatic control? Most of them tell you: "I prefer automatic control."

We trust technology. We believe it is more efficient and fairer than human decision-making. But when we asked these people: "What if the machine goes wrong?" What do you think will happen? They all agreed: "I'm lost. I won't be able to go out."

When there is what we call a false positive, when the machine is wrong, we don't have the mechanisms or the laws that guarantee that we can get out of the hole where that process of making automatic decisions has put us. Those are some of the things we have found in work in the border environment. Another example: we have also worked with universities to see the ethical impact of citizen science.

Citizen science is made with people's help. People who work on the sea movement or the migration of birds. If researchers had to take constant photographs of the sea or the birds, this would be very expensive. Thanks to technology, we can ask people who live in specific spaces to send us a picture of the sea every day at the same time. I process all this data and from there I make science. It's very good.

But when we started working with the university and the researchers, we found a paradigmatic case of what I call data despotism, which is everything for the people, with the data of the people, but without the people. We asked people, the researchers asked people to send them a lot of personal data. Don't forget that a geolocalized photograph can tell exactly where we live, what our environment is like, even who we are with. There was no precaution in relation to the management of that data.

We called it citizen science. Fortunately, today it is not like that and we have created mechanisms for citizen science to be really citizen science at the service of people and not just a space for use and abuse of the volunteers who are willing to participate in these investigations. We also work a lot in the field of work, the future of work, artificial intelligence, the robots that are going to take away our jobs, all those fears we have, things that we find.

The directors of companies, the CEOs, the managers of companies tend to make bad decisions when they get involved in technology. That is, a manager of a company

will be much more willing to spend money on technology than on non-technological processes. So we acquire technology without thinking if we really need it, especially in work environments. We are seeing the introduction of exoskeletons, 24-hour control sensors in work environments that are not necessary to improve the efficiency or profitability of that company.

But between people in companies not knowing less harmful technologies exist and technology always wants to sell you the maximum capacity it can develop, we end up with work environments that are unbearable from the perspective of those who work. We make bad decisions because technology constantly fascinates us. How do we build the mechanisms to empower everyone to know how to distinguish when a technology really helps us or not? That's what we do in the ethics of technology.

A fourth example: I work a lot in security technologies. I could give 10 TED Talks only about security because it's a terribly sensitive field. A couple of examples: a few years ago we were auditing the algorithms of a project that wanted to deploy smart video surveillance cameras in sensitive environments, banks and airports. Smart video surveillance cameras are cameras that are able to recognize

potentially criminals, automatically. It is not necessary that the human eye is controlling what happens, but they can see by the way people move if their way of moving through a space is potentially dangerous. And so, they send an alarm and human intervention is only necessary when a dangerous behavior has been identified.

What did we find? That engineers had coded themselves as normality. That is, the way to use a bank is to go in and out, because you're in a hurry, because you have work, because you're young and you can afford it. In that algorithm, those who emerged as potentially suspicious were older people, who have much more time, of course, it's raining, you go to the bank, there's a sofa, "Hey, I have a little while left, right?"

That is not potentially criminal. But the algorithm identified it as potentially criminal because it wasn't like the engineer would have used that space. In airports, families. I travel all the time, I go to the airport and I want to leave as soon as possible. But I'm not the majority. I'm not, above all, the norm. There are people who travel to take a plane once a year.

for whom the airport experience is part of the trip itself, who go five hours earlier, who go around, who eat, who eat again, who rest, who go out to the terrace. All of that is normal. It should be normal. And instead, the algorithm of the smart camera had coded it as abnormal. And that happens constantly. As engineers don't have training in that kind of thing, they don't realize that they code their normality as everyone's normality. And that, in the field of security, is problematic.

Because everything that is not normal is suspicious. And we all have the right not to be suspicious unless we do something really suspicious. So the camera constantly turns us into suspects for getting out of that normality so small defined by the technologists. Not to mention the predictive algorithms in the field of security. I don't know how to say it more clearly and more clearly.

we cannot predict the future. That is magic. And we don't know how to do it. The best we can do with algorithms is to understand what has happened so far and from there, to make certain patterns emerge and say: "The past leads me to think that it is possible that this will happen in the future." This is what we can do. Calling this prediction is very risky.

And the current insistence of the many police forces around the world to prevent the algorithm of radicalization, the algorithm of prevention, the pre-crime, is impossible. Technology can help us, but to a certain extent. And we all have the right to the presumption of innocence until we commit an illegal act.

Technology tends to forget that and feels so powerful and so under-controlled by all these elements that it dares to promise the ability to predict and know what we are going to do before we get it in our head. And finally, one last example, mobility. You've all heard about autonomous cars, right? Theoretically, in a short time, all cities will be full of cars that drive alone.

The autonomous car is one of the great examples of how the arrogance of engineering can lead to the failure of a good idea. The autonomous car is much safer than the non-autonomous car. But what happens? We are used to the deaths of non-autonomous cars. But every time an autonomous car kills or hurts someone, it is exported all over the world. So the social perception is that the autonomous car is much more unsafe than the non-autonomous car.

And while engineers insist on saying: "I have all these figures that show that it is safer", and they don't worry about convincing the population to work on the acceptability of that technology, to incorporate fears and social concerns into the conceptualization of that technology, surely what will happen is that what is a good idea that could save many lives will take much longer to implement, because as I say,

Engineers tend to be arrogant about numbers and technology. So you can see that the ethics of technology is not just a matter of your smartphone, your TV, the consumer devices we use. There is a data infrastructure that marks everything we do that urgently requires a certain control or at least the incorporation of responsibility.

What I've found in my work is that the way we make technology today is deeply irresponsible. No doctor or pharmaceutical company would ask society to launch untested drugs. Or to launch drugs that would say: "Look, this cures constipation, but it causes cancer." We wouldn't accept it. It can't be sold. And in computer engineering,

We have the right to throw things at society without testing them in any way, without having any kind of awareness of its social impact. The good news is that I think this is starting to change. There are profiles like mine, with knowledge in social science, but also in technical elements of mathematics and physics,

shows that there is an interest in these things. There are more and more research groups and work in companies, universities and public-private spaces that unite sociologists, philosophers, lawyers, physicists, mathematicians, computer engineers, to try to understand what are the legal, social and ethical impacts of technologies

that help to understand what is the desired value and how to increase the desired value of the products that are being manufactured for the future. There are more and more groups working to improve transparency and control of algorithms, for example. The idea of impact studies is increasingly widespread, in the same way that we cannot build

a new development of infrastructure, a new road, without justifying its environmental impact, we should not be able to launch technological products to the market without justifying before, that we have thought at some point, what impact will that have on society and if that impact is desirable or not desirable, and if it is not desirable, what measures have we taken to increase the desirability of that product.

Right now, many people are involved in creating the responsible innovation of the future, which I think is very necessary. And yet, we still hear voices that constantly say that prevention, ethics and regulation go against innovation, that we cannot limit the creative capacity of men engineers and their great ideas. Well, I'm sorry, but that's not the case.

As humanity, we have been carrying our entire history, domesticating technology. Does a car need a safety belt to circulate? No. Does a car need a speed limit? No. Cars could go much faster than they do on our streets. Does a car need zebra crossing, traffic lights, pollution control? No. We have added all these things

human societies, to ensure that what is a desirable technological innovation, which is a motor, something that takes us from A to B faster, is also desirable from a social perspective. We have limited the possibilities of cars, of vehicles, because we have understood that we want to go faster from A to B, but we don't want so many people to die in that process.

We have always done that. We have always limited the capabilities of technology because technology only makes sense if it improves, if it addresses social issues. That computer engineering has forgotten that logic so many times is clearly worrying.

In the end, what we have done with the car is to have debates over a long period of time. The first car did not have a seat belt, nor did there were traffic lights, nor did there were zebra steps. But for a long time, as societies, we created these frameworks, these social consensus that lead us to be able to use vehicles in a normalized way.

These are the debates, the social consensus that must emerge around any new technology. We must understand that when a new algorithm, a new technological innovation or a new invention appears, we must decide among all of us which will be the seat belts, the speed limits, the emission limits, the zebra steps and the traffic lights of that new innovation. This is what the ethics of technology is dedicated to.

First, to encourage these debates, so that we all decide whether it is better to have human control or mechanical control when crossing a border, so that we all take responsibility for the decisions we make in relation to technology. And from that desire, what we see is how to build fundamental values, social cohesion, trust, non-discrimination, human rights, in the technical specifications. Because it is not just a matter of giving speeches. There are ways

through technical specifications to make technologies respond much better to society's expectations and wishes. In the same way we have been able to build cars

that kill less. We can build technologies that are not only less harmful, but that contribute to improve the future of our societies. In short, the ethics of technology is to put people at the center of technological development and to put technology at the service of people, and not as we are now many times, which is people at the service of technology. Thank you.