Technology is reshaping the world of work with automation, artificial intelligence and robotics increasingly being integrated into workplaces. What advantages does the so-called Fourth Industrial Revolution present for women?
There are two key ones I would like to highlight. The first is that, research has shown that some of the careers that won’t be as impacted by the Fourth Industrial Revolution, are the caring and teaching professions, professions where you have to connect to human beings. If this is so, there’s actually a huge opportunity for women who work in these industries. They almost have a head start over men since they have the opportunity to keep their jobs in these sectors, move up the career ladder and attain managerial positions, conditional on their expertise being valued in these professions of course.
The second is the set of opportunities that technological change presents for women who work in industries where you don’t have to be on site. With a sector like mining, you have to be there physically working in the mines. But a lot of the technical roles will be able to be done from a distance. So there is the benefit for women to be able to learn, train and work at a distance in these roles, working flexibly, even more so then they are able to do at the moment.
The technological transition has also been criticized for the negative impact it could have on workers, who could increasingly be replaced by these technologies, with women expected to be more disproportionately affected than men.
What are the challenges that women face in this time of change and how could it affect women differently in developed and developing countries, in low-skilled and high-skilled jobs and from low-income and high-income backgrounds?
Currently, we don’t value the work that women do, which if everyone started to fight over [the same jobs], women would face the risk of losing out over their male counterparts. The reason for this is, if something is seen as important, the men usually take over, which is what we’ve seen happen in the technology industry. Originally, it was women who were doing it, and then when it became more important, and more commercially viable, we saw the men get involved. So what we’re seeing is that, when we talk about people losing work, it’s usually the women that are giving up working to do childcare – and everything else that counts as unpaid labour – at home.
The other problem is with the biases we’re seeing being replicated in some applications. When things are replaced by algorithms that are trained on historical datasets, we’re seeing these applications making decisions based on biases against women. Whether that’s in recruitment for jobs, decisions being made for mortgages and everything else you can think of where a computer can literally say no to someone because of the biases that have been fed into them. Without us being proactive in forcing people to be, not only ethical, but equitable, in what they’re building, we will likely continue to see women being disproportionately affected.
How could it affect women differently? I’m not sure. The fact that we are much more globally connected now than in the past few technological revolutions means that, anything that happens to women in the developed world, will happen even more so to women in developing countries.
If we look at the number of women working in the science, technology, engineering and mathematics industries at the moment, we know in developing countries, the proportion is slightly higher, so that means they have more technical women working in these sectors, although whether they are staying in those sectors is another matter but, importantly, they’re doing better than women in developed countries.
But technology is only as good as the adoption you have from the people and the reaction to technology has been different around the world. That might have something to do with language: a lot of these programme languages are in English so there might be a correlation between English-speaking versus non-English speaking countries.
The impact of artificial intelligence, for example, is also shown to have more of an impact on women in developed countries than women in developing countries because of how much it has been realized in the developed world.
Research, as you mentioned, has revealed that some applications based on machine learning have been found to replicate the social biases which are fed into them, notably biases around gender. How can these processes be improved, so as not to perpetuate gender biases, and who is responsible for its oversight?
That’s the big question at the moment. Should it be governments or technology companies or someone else overseeing the technological change that the world is experiencing? There are several quick wins we could do to improve these processes. For example, regulation can help – although it feels like a stick rather than a carrot in trying to solve the problem – but it is one that needs to be considered because the sector doesn’t necessarily see it as its responsibility. For example, recently, Mark Zuckerberg said that governments should be regulating the public’s use of data, which is a bit of a strange thing to say, when he is the one who has created the platform that’s using our data.
Interestingly, the Institute for the Future of Work is looking at frameworks you might be able to apply to the use of algorithms at work to ensure you have considered what ethical practice looks like and any gaps in the datasets that you have before building anything. It’s important that we don’t just have people who are technically competent working in roles but that they also have some sort of training in ethics. This will help to limit the probability of any biases being fed into the codes when building applications.
In terms of whose responsibility it is, I think it’s a joint responsibility, because it’s a joint risk, since all of us lose out when our applications are biased. Governments lose out, companies lose out and society loses out too. That’s why I’m advocating to set up a discussion across society where governments, companies and the layperson can come together to discuss the best way forward.
We currently have this thing in technology where it’s the programmer against the user when we should all be participants in the system. What this means is that, for better or for worse, everyone should be able to contribute to the discussion around technology in order to arrive at appropriate solutions. But this involves a shift of power which we know is a difficult thing to get people to do. But it is an important step that we have to take otherwise we will get companies marking their own homework which is not good.
