The future of emerging technologies depends on getting the political and social context right

Do you think there is a lack of understanding of digitalization within the development community?

I think there are two sides to it. There are many organizations interested in applying emerging technologies in internal back office functions because it will help them to be more effective and efficient. It is a safe thing to do and you don’t have to worry as much about the external environment. What is going to make a bigger difference is work focused on the enabling environment. There are huge gains to be had by helping developing countries grow their own digital ecosystems, rather than coming in with high tech interventions that see that technology as something that belongs to us. I think we are a lot better off helping people get access and then letting them use it to solve their own problems.

Poor people often lack access to digital tools. There is a need for infrastructure. Many emerging technologies only work on 4G or 5G networks and many developing countries still work with 2G networks. Do you think the digital divide will increase with these emerging technologies?

 I think a lot of the real cutting-edge technology development is being done by private companies, because that is what private firms do. And these companies can maximize profits by focusing on high-paying customers in the developed world. This can lead to a digital divide. That is why it is necessary for development agencies like USAID to be engaged with this digital development space. We are helping to bring connectivity to places where it might be too risky or where it might not be sufficiently profitable right now, but where there is potential for growth in the future.

Which technologies are currently applicable in the development context in an effective way? Which ones are, maybe, not appropriate?

I am always reluctant to say, ´This technology is never going to be applicable’, or, ´Yes, this technology is going to take over the world´, because it is easy to be proven wrong. Rather than looking at a specific technology in isolation, we like to think about what we call sociotechnical systems, which are a combination of people, institutions, technologies and policies, that get things done in this system. The first thing we looked at when we launched our team was Digital ID, which is not a single technology. It is a whole set of technologies that includes well established things like smart cards or biometrics and some things that are a bit further out there, like blockchain, for example. But they all come together in this space that also involves a lot of politics and thorny user design questions. Whether a given technology is going to be applicable or useful depends on getting that entire political and social context right.

Do you have an example of a country that applied that sociotechnical system well and implemented an emerging technology recently?

My favourite example, because it is one that we can learn a lot from, is the Aadhaar ID system in India. They decided several years ago that they wanted to issue a unique ID to every person in India, which was quite a task because it´s a very large country. They used biometrics and a standards-driven design approach, where, instead of the government going out and procuring something, it set standards for data quality and let the private sector compete to offer the services that would work. There were many things that they did well: they were able to bundle identification to enable access to financial services, e-government systems, and to unlock a whole lot of other modern services that otherwise would have been very difficult to do in India. There were also some problems with Aadhaar, and we can learn from those, too. When thinking through privacy, the people who designed the system paid a lot of attention to how they thought it was going to be used. They ended up being very surprised by some of the ways in which it was really used, which have eroded some of those privacy protections. There have been some prominent data breaches that made people very upset. Ultimately, the Supreme Court ruled that the government of India could not make enrolment in an ID system mandatory. It has been a good example that we can all learn a lot from.

Do you have an example where it did not work that well?

It is important to make sure that you are keeping the problem in mind from the outset and not letting a desire to use a certain technology drive your solutions. In the blockchain space in particular, a lot of things get rolled out where a centralised database may work just as well. One can say it’s not necessary to use this complicated cryptography-backed blockchain system in order to perform the task. That is something that may make people think that it´s all a hype.

Could you give some insights on where blockchain could be useful? Do you have any advice for people who want to start working with blockchain but don’t know much about it yet?

Often, it is about situations where you need to have a permanent shared record that is accessible and writable by a large number of people who do not have complete trust in each other, or at least there isn´t one party whom they can all trust. If there is one group that everybody trusts, you can just ask them to maintain a database and there is no need for this complicated blockchain system. One example that people are talking about is supply chains, where you have different entities. They might be companies, some of them being partners, some of them competitors. They all have access to different information. In general, they don’t want to disclose all of their internal business processes to some external database manager. If you use blockchain in a supply chain to track the movement or the handoff of goods from one individual or organization to another, that might be a good application for it, in some contexts.

If you were a social entrepreneur in a developing country, who did not know much about these technologies, how could you learn about them?

A report that the U.S. Global Development Lab recently came out with is on Artificial Intelligence (AI) and machine learning. It contains two in-depth case studies. One of them is with a company in South Africa that is the kind of social enterprise we are talking about. The problem that they were tackling is youth unemployment.. This company, Harambee, is looking at trying to match youth who are at risk of long-term unemployment with their first job. They are looking into AI and machine learning as a way to help them reach larger scale and be able to match people more accurately to a job that fits their personality and skills.

Another group that we looked at is CIAT, the International Centre for Tropical Agriculture, in Colombia. They have been working with farmers´ cooperatives and using machine learning methods to analyse the data that they have gotten from these associations to be able to provide the farmers with science-driven information and predictions that will help them optimise their yields and respond better to changes in climate.

In both cases, they try to integrate their tech people with their non-tech people, or, to put it in another way, to integrate their solution people with their problem people. That can be really hard because of different cultures, styles of working, and priorities, but if you can get that right, it can be powerful.

Related publications:

• Primer on Blockchain
• Reflecting the Past, Shaping the Future: Making AI Work for International Development
• Identity in a Digital Age: Infrastructure for Inclusive Development