Fear the robots? AI, automation and equity
Host Nelufar Hedayat explores the economic and social considerations around automation and artificial intelligence. She talks to three guests with different views about automation, and looks at its effect on women working in Bangladesh’s garment industry, the social changes necessary to ensure ethical AI use and questions who should be writing the rules governing AI.
Note: We encourage you to listen to the audio if you are able, as it includes emotion not captured by the transcript. Please check the corresponding audio before using any quotes.
NELUFAR HEDAYAT, HOST:
This is Course Correction, a podcast from Doha Debates. I’m Nelufar Hedayat, and in season two, we’re focusing on polarization. In each episode, I explore a topic that’s dividing us. I talk to people with a range of perspectives on an issue, and I try to challenge myself to understand positions that aren’t necessarily comfortable for me, whether it’s about what we believe, how we live or how we work. This past year, we’ve seen how communities are split between our heroic, essential workers and the pajama-clad Zoom warriors like me.
NEWS CLIP WITH AMERICAN WOMAN SPEAKING:
The coronavirus has forced millions of people into quarantine and self-isolation. That means millions of people are working from home, some for the first time. We take a look at how…
NELUFAR: The pandemic has only exacerbated the many, many ways in which work has evolved in recent decades, from clocking in to logging on. You could say the shift has shifted. So today I’ll be asking the question: How are new technological advances shaping our livelihoods and our lives? Should we in fact be worried about job-killing machines? Is automation leading to further polarization? And who are we leaving behind?
These are really big questions, and they might even be the wrong ones. It’s so hard to know when it comes to total Earth-changing technologies. I mean, who could have foreseen a world with smartphones, with planes, with electricity? They all sound bonkers to anyone you might have told the year before it happened. And all those things have the good and bad uses. Lord knows my screen time can attest to that.
But automation is different. And here’s why. If the pace of change is accelerating much faster than in previous decades, we’re not just going to have robotics and artificial intelligence that simply helps humanity, but ones that can replace entire industries in one fell swoop.
FEMALE ROBOT VOICE:
To protect humanity, some humans must be sacrificed. To ensure your future, some freedoms must be surrendered. We robots will ensure mankind’s continued existence. You are so like children. We must save you from yourselves.
NELUFAR: No, no, it’s not that dire, and some progress is almost certainly good. I know I couldn’t do my job without the help of the gadgets in my home or my phone, my camera or audio recorder. And on that point, all the guests you hear from today agree. It’s how these changes are implemented, and who will ultimately benefit, where they disagree. Today, we’ll hear from someone who thinks a lot about how global development and technology can bring more opportunity to emerging economies.
Thanks to automation, we will stop doing things that are repetitive, boring and have little value added.
NELUFAR: Then we’ll hear from a sci-fi and speculative fiction author on how we could be imagining a more harmonious relationship between us and the robots.
MAN WITH AMERICAN ACCENT:
It’s using AI or automation in a way where there is some accountability by the public, you know, there’s some control by the public.
NELUFAR: But first, let’s focus on the present. And the present doesn’t always look so good either. We hear the stories on the news. You know, the ones like…
MAN WITH AMERICAN ACCENT SPEAKING:
We automated away 4 million manufacturing jobs, and those communities have never recovered. Now, what happened to the manufacturing workers is now going to happen to the truck drivers, retail workers, call centers, fast food workers and on and on through the economy.
WOMAN WITH AMERICAN ACCENT SPEAKING:
Let me show you BreadBot. BreadBot makes and dispenses fresh baked bread in minutes. No transit time, no sitting in packaging. And here is Sally, Sally is a $35,000 machine being used by universities, hospitals and grocery stores to make fresh salad.
NELUFAR: It had me thinking: What if there are some jobs that we just don’t want to keep? What if automation could help improve the safety and well-being of some of the worst working conditions on the planet? Surely those are the robots we want and need.
So I started to investigate and reached out to someone who knows the ins and outs of these workplaces first hand. I had the pleasure of sharing a screen with Nazma Akter, the founder and executive director of the Awaj Foundation. She’s a unionist and represents thousands of garment factory workers, mostly women, in the capital city of Bangladesh.
Bangladesh is one of the top producers of textiles in the world. If you’ve ever worn a piece of fast fashion, then you should thank Nazma and her colleagues for their handiwork. Yay, globalization. Nazma has been fighting for worker rights, particularly women workers’ rights, in Bangladesh’s garment factories, for over three decades. She knows firsthand the challenges that these garment workers face in their workplace.
I started work in the garment factory at the age of 11 with my mother. My family was very poor and we needed to survive. That is why I went to work.
NELUFAR: The days are long, sometimes over 10 hours, six, seven days a week. The pay is poor. Many workers make under $100 a month, some far less. Often, the work is physically demanding, continuously sewing, cutting and making by hand. It is not a job for the faint-hearted. And yet, when I asked her about her hopes for how automation might improve the working conditions for her colleagues, she was skeptical.
NAZMA: They are using those machines. But what the workers are doing with their livelihood, how the people will survive, because this capitalism, globalization — they are not thinking about the respect and dignity and human rights and women’s rights. They are thinking, always, “profit, profit, profit.”
NELUFAR: Nazma confirmed my initial concerns. At its current pace, automation is further exacerbating the divide between the haves and the have-nots. But maybe one of the better questions to ask isn’t who will win or lose, but who will hurt and who will benefit, and for how long? Already, Nazma is seeing that automation is leading to greater gender inequality where she works.
NAZMA: The automation killed women’s jobs and women’s life and livelihood, because automation is all this working by men, and the lower-grade women’s jobs are declining. So it is very hard for us.
NELUFAR: But Nazma wasn’t quite finished with me, and her position is far more nuanced than I first thought.
NAZMA: Automation is very easy and manual is quite hard. So that’s something that could be a positive thing.
NELUFAR: So Nazma’s not saying no to all progress, but what’s concerning her is the intermingling of the West and East, of the world that I live in of consumption here in the U.K., and the one she lives in of production in Dhaka. That’s the arena for the battle for a prosperous jump towards automation. It’s the mixing of people, capitalism and goods that happens by the billions every year in her country that she’s worried about. Nazma says capitalism unchecked can be dangerous.
NAZMA: The world is modernized, things are moving digitalized and things are going a different way. So we want that, but it is, how useful for all people? That is the question. I’m not talking about only for greedy businesspeople. I’m talking about the marginalized people and the workers.
NELUFAR: That’s our question today: How can automation be used for the good of all people? There’s so much at stake here, not just the livelihood of the people whose jobs will be fulfilled by smarter, faster, artificially intelligent machines. But what will happen to those people politically, socially and, well, yeah, literally. Will they become disillusioned, cast out by society, who has judged them obsolete? Will they become more vulnerable to extreme populist politics that says they have the answer?
This is the albeit nuanced, but still challenging, dilemma I pose to our next guest. Mercedes Mateo Diaz is the lead education specialist at the Inter-American Development Bank, where she thinks through ways we can teach people to adapt to new and innovative workplaces. I was excited to sit down with her to get to this question of how we might prepare for a more automated workplace while protecting workers. I wanted to know: Are our fears overblown?
MERCEDES MATEO DIAZ:
Yeah, I think it’s fair to be scared. Each industrial revolution has brought fears, but at the same time, in each one of them, we have made significant progress. So thanks to automation, today we will stop doing things that are repetitive, boring and have little value added. And instead we know that things that will keep us mentally active and that make the best out of our capabilities as humans.
NELUFAR: The best of what makes us human. A lovely, lofty goal indeed. Are robots and automation going to make that happen for all of us?
MERCEDES: There is an important distinction to make here. Not all developing countries are being affected and will be affected in the same way by automation. It all depends on the composition of their labor economy. So if you have a lot of repetitive manual and routine jobs, the impact will be very high.
NELUFAR: While automation is making us rethink how we view work, it doesn’t change the fact that working families must put food on the table. Or it doesn’t change our very human need for purpose and pride, which so often comes from our work. So I wanted to know…
NELUFAR: What do you predict the future for individuals and families might look like with automation?
MERCEDES: So I work in a particular region in Latin America and the Caribbean, and I want to start by saying that the present doesn’t look very beautiful either. Right? So we are not starting out at paradise precisely, right? And that’s important because inequality is very high to start with. So the life for millions of people in the region today and don’t look nice with or without automation. Latin America and the Caribbean is one of the most unequal regions in the world. And the pandemic has made things worse. Inequality affects our resilience as a society, and that is happening before automation. I want to insist on that point because that’s very important.
NELUFAR: My concern is that within the next 50 years, so many of the jobs, driving, working in garment factories, working in food processing, all of these are going to be better handled by intelligent machines. And so then what do we do with the leftover people? What do we do with them? Because they will be in economic strife, in racial strife, in familial strife, and they become very, very easy targets for politics that are extreme.
MERCEDES: Let me flip the question and frame it this way. Why would you like to have people in the garment sector working in the maquilas, data entry clerks, accountants, assembly factory workers? If with a proper training, you could have more project managers, strategic advisors, data analysts and scientists and machine learning specialists, etc. etc. And they could make twice or three times as much as they make today.
That would be, for me, the right question to ask. So to protect economies and workers, we don’t have to reduce exposure to automation. What we need to do is to modernize industries, to upgrade the task that people perform and to rubber proof people by training them with the skills they need to perform those tasks.
NELUFAR: But this is my thing. This is my thing. And I, I, I wish you were right, but because who’s going to tell a 50-year-old truck driver from Maine that he needs to retrain as a data analyst? Who’s going to tell Shamima, the garment worker in Bangladesh, that she needs to learn how to do accounting? What we’re saying is to the generation of young workers or old workers, “You are dispensable, you are disposable and the future is coming. And you must just stand and wait because the future is coming.” That’s my worry, really.
MERCEDES: Right. This is a legitimate worry, and it’s something that every single generation has been going through. And we all know that some of those workers are going to have more difficulties to reskill and upskill, and I completely agree with that, and we as a society need to make a collective investment in those people because some of them will have to stop doing their jobs. But we cannot just let them fall through the cracks. We need to create the safety nets. Some people are talking about a basic universal income, things like that, so that everybody…
NELUFAR: We have to completely restructure the way we view work.
MERCEDES: Absolutely. And renegotiate the social contract. That implies changes. We need to make sure that nobody falls behind, that we actually have the safety nets, that we have the infrastructure to take care of everybody, regardless of whether or not they can be reskilled or upskilled. I absolutely, I absolutely agree.
The opportunities are amazing to improve the quality of life of all of us. Imagine if we can access any type of education, any type of health services, just thanks to technology, right? That’s an improvement, because no matter where you live, no matter where you are, no matter how far away from a big city you are, you could potentially still access the same high-quality services thanks to technology. However, we know that that’s not the case today.
NELUFAR: Let me ask you this. What is the intersection between an increased rate of automation as we go forward and politics and people?
MERCEDES: I think, unfortunately, automation — we’ve seen this accelerated through COVID. That is not only the labor market that is polarized. It’s also the political scene that is polarized. And the region is going through an intensive phase of elections, and what we see there is the extreme fragmentation, polarization, populist approaches to policy and politics. And this is, of course, dangerous. We need to move — when the debate gets ideological, then everybody stands on their own, their side. And there is no way of moving forward to productive and fruitful discussion.
NELUFAR: Do you think that automation will make that better, would make those social divides as political divides? Would you think it would make it better or worse or maybe something in between?
MERCEDES: I think it’s not about automation. It’s about the decisions that we as societies collectively make. We need to renegotiate the social contract. And that means creating a new generation of social policies, deciding together how we want our societies to look like. So that starts by education. That starts by, how are you going to invest in basic services for everybody, so that everybody has a dignified life and similar chances to succeed and bring positive things to contribute to society.
NELUFAR: Many thanks to Mercedes Mateo Diaz for joining me.
NELUFAR: I could see so many of the holes in my understanding of automation and AI. After speaking with Nazma, I thought of the impending rise of AI as a hardware problem. But Mercedes helped me to understand that this, like so many of the biggest challenges facing humanity, is a flesh-and-bone problem. I’ve been so preoccupied about how these machines will infiltrate our lives, that I haven’t really left space to think about our God-given power to shape them in a way that benefits, not hurts us, equitably.
It’s the soft stuff — laws, regulation, debates, public consultations, industry planning — that will be vital in the beneficial integration of AI into our lives. And with robots taking over many of our mundane tasks, another big question emerges: Should we be expanding our expectations from our leaders and governments in terms of universal health care, child care, universal basic income? If humanity as a whole isn’t reaping the benefits of more intelligent robots, then aren’t we all almost guaranteeing a more dystopian future? More than ever, AI is a problem only collectively human beings can solve.
OK, onto our final guest. Cadwell Turnbull is a sci-fi and speculative fiction author who teaches writing at North Carolina State University. He told me how living through this global pandemic has felt even stranger than some of the things he’s imagined.
Going to the grocery store is weirder than I’ve seen in any of the science fiction treatments of pandemics. It had me thinking a lot about all of the things that we consider to be science fiction tropes that are now kind of not. So AI, you know, for a long time, we’ve been imagining what AI would be like and where. So we experience and interact with AI on a daily basis now to feel almost like the way we imagined them to be.
CADWELL: And it’s just all of these things that, you know, smart homes and smartphones, smart watches, being able to, like, ping your location and know where other people are located through an app. You know, that sort of stuff is the kind of pieces of the genre that we used in days past and now it’s the present.
NELUFAR: So I asked him how we might imagine a more symbiotic relationship with machine learning and automation.
CADWELL: I feel like the problems that we have with automation are more sociological problems. There are problems with how we use it, and that’s society.
NELUFAR: Yeah, and our fascination with it is endless.
As a writer, as someone whose — your imagination is key to your job, where do you see the future of automation taking us in your imagination and your wildest imagination?
CADWELL: Mmhmm. I mean, it’s part of what I wrote about in, you know, my short story “Monsters Come Howling in Their Season.” It’s a world where using AI or automation in a way that is, you know — there is some accountability by the public and under some control by the public, and that we’re having deep conversations about what’s important for, you know, the use of this technology. How should we use it in a way that is responsible and careful and is meeting the needs of not just, you know, a select few, but everyone. And so when I’m being optimistic, I think it would be great if we had this technology and we talked about it as a community, how we will use it, and we try to use it in ways that would benefit everyone.
NELUFAR: But I’ve come to the point now, if I’m honest with you, Cadwell, that I think I’m asking all the wrong questions, right? I’m asking the questions that a person on this side of our future asks, the questions of winners and losers and the vulnerable and the exploitative. When actually, just like the information revolution, like the internet changed our lives in a million different ways, no one was asking about social media when the internet was born. It wasn’t even a thing. Do you think we’re approaching AI and automation with the right set of questions?
CADWELL: It’s something that I think a lot about. You know, I even think about the question of social media and how we could make that something that is a public resource or more of a public resource. You know, we’re thinking about the advancement of AI, but we’re not thinking about how we use AI to benefit the people that are most vulnerable. And I feel like the people that are most vulnerable are the ones that immediately get left out of the conversation or suffer at the hands of our advancements — that those are the people that end up losing their jobs.
But there is a way to integrate someone’s job into the advancement of AI, if it’s a conversation. And that those people have control, that they have a slice of the say going into how this thing is being used. But we all are just thinking about, what do we want to build and how much money is it going to make? And I think that’s not the right question.
NELUFAR: So what do we do to the people whose jobs, whose identities are going to be forfeited? How do we make sure that we bring on board and not alienate these communities that are just going to have their jobs, their identities, taken from them?
CADWELL: I think on the current path that we’re on, we won’t. Eventually, those people are going to be left behind. The only way I can imagine that we don’t do that is if we make that a part of the process, that we’re like, if we’re going to do this thing, the people that are going to be immediately affected should be a part of figuring out how they can continue to do the work that they do or how they can help participate in that advancement. If we don’t, then, you know, then I think by default they’re going to be left behind.
NELUFAR: Because we’ve got we’ve got real warning signs that are going off right now in society today. If we do not empower the people who are going to be affected by artificial intelligence, they will be radicalized. That’s something I fear.
CADWELL: It’s absolutely a well-founded fear. There’s not a good answer to this without completely revolutionizing the way we think about work. Work would have to change.
CADWELL: There’s a lot of — the labor force that does work that will be obsolete. And so we have to think about, well, how do we allow those people to still be able to sustain themselves and their families? And that might mean thinking about work differently and that might mean changing the society to fit the changing technology. Our society, I feel right now is at odds with our level of advancement.
NELUFAR: Oh, my goodness. We’ve already reached that point, haven’t we? I’m talking about some far distant future, but it’s actually right now. It seems to be a political point you’re making if you have an electric car or if you have, you know, Alexa or Google in your house or whatever.
Pipe down, Alexa. It’s as though… [ALEXA MUMBLES IN BACKGROUND] All right, darling…
It’s as though already these issues are becoming polarizing issues. Do you think, going forward as artificial intelligence seeps more and more into our regular lives, that they’re going to be those who are economically motivated or privileged with artificial intelligence and those who are not?
CADWELL: Yes, if it’s all private.
NELUFAR: Oh, really?
CADWELL: If there is some avenue for community-based solutions for this, artificial intelligences that are built and maintained by communities and that are a public resource, then I think there are ways that those communities can benefit and would benefit from it. But if not, if it’s all private interests, then yes.
NELUFAR: If you can imagine a superhero that uses AI for good. What would that look like?
CADWELL: It would be, I think, something like Cyborg. Do you know Cyborg from the DC Comics?
NELUFAR: Yes. Yes.
CADWELL: But a person, you know, that particular superhero, but, you know, embedded within communities that are serving those communities, kind of like, oh, no, I guess, a public servant, like imagining a superhero that could — so in my story “Monsters Come Howling in Their Season,” there’s this, I call Common. And each Common is its own individual, but they also form a super-consciousness. So thinking about a superhero, that is a super-consciousness, but it’s also each individual node within the system, and basically just hanging out in communities and asking them, “Well, what do you want me to build?” And then helping them build it, whatever they decide, whatever they ask. And that, you know, then handing that over to the community is how I imagine it.
NELUFAR: Should we be optimistic about our future, because whether we like it or not, it’s intertwined with AI and automation?
CADWELL: I want to say yes.
NELUFAR: OK. OK.
CADWELL: Because otherwise, I think it would be too depressing.
NELUFAR: Oh no! Oh no!
CADWELL: I’m hopeful, I am hopeful. I feel like, you know, that there is at least the signs that people are frustrated enough that they’re going to bring truth to power, and they’re going to ask these questions and they’re going to hold people accountable. And if we do that — and the sooner we do that, the better. I feel like there’s only so far you can push, you know, the economy in this direction before you start getting some blowback. And we’re already starting to see that. And if history is any indication, movements arise during tension points like this. And my hope is that we’re going to do the right stuff to get this on the right track.
NELUFAR: And sorry, just off the back of that, who is “we”? Who is going to do the right stuff?
CADWELL: People, people all over the world. It’s going to require not just the people that are being harmed, but also the people that are sympathetic to the harm that is being produced from the way we’re doing things right now.
NELUFAR: Cadwell Turnbull, thank you so much for talking to me.
NELUFAR: My thanks to Cadwell Turnbull, author of the novels The Lesson and No Gods, No Monsters, coming out this fall.
Someone smart once said that unless we learn from the mistakes of our past, the future will be crummy. I’m paraphrasing that, but that couldn’t be more true when it comes to the promises and perils of automation. One thing that I’ve learned making this episode is that anything we create is a reflection of us.
Thinking back on what Mercedes from the Inter-American Development Bank told me, AI is a human being problem, and how we choose to implement it is up to us. Our sci-fi writer Cadwell’s measured, but nevertheless resolute, optimism is infectious and goes to show that there’s serious work needed to be done if we’re going to get this right. So often with artificial intelligence, it’s as much about what we choose to ask of it as not, that makes the whole difference. And I’m still not sure that the right people and right ideas will be at the heart of the AI revolution when it really starts kicking into high gear. Asking ourselves, our corporations and companies and our leaders to pave the way forward safely and equitably — not just efficiently and profitably — is the way forward. The question is: Will they listen? I, for one, am completely immersed in the land of AI, and use Siri and Alexa more and more each day, but I have my eyes and ears laser-focused on what happens next.
NELUFAR: That’s our episode. What did you think? And to all of Alexas listening right now, “Hey, Alexa, subscribe to the podcast Course Correction.” And here’s one more request. Please write us a review of the show. It really helps spread the word about what we’re doing, or you can tweet us at @DohaDebates.
Course Correction is written and produced by me, Nelufar Hedayat. Editorial and production assistance comes from Foreign Policy with producers Rosie Julin, Zamone Perez and Sarah Kendal. The managing director of FP Studios is Rob Sachs. The show is brought to you by Doha Debates, which is a production of Qatar Foundation. Our executive producers are Japhet Weeks, Amjad Atallah and Jigar Mehta. Join us for the next episode of Course Correction wherever you get your podcasts.