Pathways to Impact: Claire Melamed

Pathways to Impact is a series of conversations with data for social impact leaders exploring their career journeys. Perry Hewitt, Chief Strategy Officer of data.org, spoke with Claire Melamed, the Vice President of AI and Digital Cooperation at the United Nations Foundation.

How did you come to do social sector work?

I grew up in a household that made me very aware of social and political issues. My parents were very political; they lived in the US in the 60s and had been involved in the anti-Vietnam movement, and in the solidarity movement supporting the anti-colonial struggle in Mozambique. We all moved to Mozambique as a family after independence, where my parents worked in the Ministry of Health. It was actually more of an assumption that the purpose of working was to try to make things better in the world.

When was there a pivot toward data in the work that you do?

My first degree was in politics and history, social science, and not really hard sciences or data. In my 20s, I did a PhD and went to live in a small village in the north of Mozambique for a year to do research on cotton farming.

There was very little data about the village. There was a census that had been done some years back, and I decided, because that was the standard methodology at the time, to do a couple of rounds of a small survey. My hope was to understand some of the broader dynamics of the village and how cotton farming was shaping the economy.

The data revealed patterns that I would not otherwise have seen about the way that people were living. For example, it showed how men and women farmed differently, with men mainly farming cotton and women mainly farming maize and other food crops. In each scenario, they were employing different kinds of labor on their fields, and paying them in several different ways: cash and maize as an alternative sort of cash. The outcome of the survey data was that it really changed the trajectory of my dissertation, the nature of the subject I was exploring.

This early interaction with data was one of those formative experiences that shifts quite fundamentally the way you see the world. The data showed how the community’s experiences were quite different from how I’d understood them. The combination of the two- having new data and also living embedded in the community- changed my understanding of why people were making the decisions they were making and what the impact was. That was my first experience of the power of having the numbers as well as the stories, a combination that can offer you a more rounded picture of what’s important to people.

I think about data as a problem-solving tool in two ways: the power of data to understand people's lives and the power of data in politics, where data is a tool to influence change.
Dr. Claire Melamed Vice President, AI and Digital Cooperation Strategy United Nations Foundation

How do you see the impact—realized and potential—for data and AI in solving social problems?

I think about data as a problem-solving tool in two ways: the power of data to understand people’s lives and the power of data in politics, where data is a tool to influence change. Working in non-profit and campaigning roles and then working with the UN as part of the negotiations around the Sustainable Development Goals gave me a different understanding of the power of numbers. What eventually brought me to the Global Partnership for Sustainable Development Data was this belief in data’s power to both understand people’s lives and to shape politics. Because data is so powerful, the data systems that underpin understanding and political decision-making must use the best available technologies and methodologies—we owe it to the people whose lives will be shaped by the data that’s collected and used.

The need to invest in data for government decision-making is not new: the first census was done in Babylonian times.

What’s new over the last few years is the role of data as an input into AI systems, and that gives us an entirely new dimension to this conversation. Now data is not only an input into human understanding and decision-making, but it’s also an input in technologies that produce knowledge in a new way. Getting the data right is a critical investment in the understanding that we’re building into machines, into AI, and our infrastructure. All this has massively increased the value that we place on data and the importance of data in shaping our lives.

We’re very much at the beginning of this story, so we don’t know how it’s going to end. But I do think there are a few constants that we need to think about, one of which is the need for skilling and infrastructure to make sure that this new reality can be shaped by everybody.

There are real concerns about the level of concentration in the AI industry in general. There are a few companies that are determining how products are developing, and even what the public knows and thinks about this technology, given the extent to which they’re dominating the media narratives around AI. If the AI industry is to evolve in a way that benefits us all, then more people have to be much more involved in the development of that industry. That comes through people having the skills, knowledge, and confidence to be involved. It also requires living in countries where the digital infrastructure allows them to do that, from connectivity to hardware—that means all countries should have that essential infrastructure. There are a lot of different things that need to happen to make this data and AI revolution inclusive.

Part of the Pathways to Impact series

Curated conversations with data and AI for social impact leaders on their career journeys

See all Pathways to Impact

Which skills, not necessarily related to data and AI, have helped you in your career?

I’ve worked for a long time now in a field where people sometimes feel excluded because of their lack of technical skills. So, I’ve heard again and again people saying to me, “I don’t really know enough,” or “I’m not a data scientist, how can I have an opinion on AI?” And every interaction in every job has convinced me that the most essential skills are not the technical ones. I am pretty numerate, but I have no technical data science skills; I’m not a programmer. But you don’t need technical skills to have a point of view about how technology impacts society; you just have to live in society. It’s really important that everybody has the confidence to feel that they can play a part in this data and AI arena, and they do not exclude themselves because of their lack of specific technical knowledge.

The critical skills are the ability to interact with people and to have the empathy and humility to be able to listen to people. It’s a skill to properly listen to what people are saying without imposing your own worldview, and instead finding points of connection and consensus. Honestly, I think these skills are the most critical for navigating this particular moment in time.

What are some of the obstacles you’ve overcome?

There are internal and external obstacles. We were just discussing some of the internal obstacles that get in our way—imposter syndrome is something almost everyone experiences.

The roles I have held involved a lot of coalition-building and convening. And the two big blockers that we all face, to different degrees and in different combinations, are political and financial. Is the political environment conducive to what you want to do? And are there enough resources? These have been recurring challenges throughout my career.

A big challenge I have seen repeatedly is starting with a very big ambition of some massive thing you want to change, and realizing that while you want to change this big thing, you need to figure out how to break that down into smaller things that can happen in the meantime. It’s the art of trying to develop a strategy by identifying the first step. That step is often much smaller than you want it to be because of a political landscape that is hostile or just not ready.

The second obstacle is financial. This is particularly acute right now. And even though we think of this issue as primarily affecting nonprofits, if you are inside a government or even a funding organization, your team has to get resources for the things that you want to fund, and you’re negotiating with your colleagues on that. Everybody has to think about resources all the time, just in a slightly different context. In any sector, you have to persuade other people that what you want to do is important and necessary, and more so than many other important initiatives or opportunities that also exist in the world. And that the plan that you have to effect change is feasible and worth investing in.

It goes back to the non-data and AI skills that matter: listening to other people, empathy, and finding common ground. To overcome the common obstacles of politics and finance, you need those listening skills to understand where people are coming from and, ultimately, to persuade stakeholders.

What community of people or resources bolsters your work, at the UN Foundation or throughout your career?

What’s so fantastic about the UN Foundation, and any organization whose purpose is convening and networking and finding unexpected coalitions, is that it gives you broad exposure to colleagues from so many different communities.

My jobs, both in this role and back at the Global Partnership, involved finding the people from any sector who want to get down to business. People who share the same values, aspirations, and goals. The shared challenge becomes how we can find ways to work together; how we can bring the best of all the different communities, the incentives they are facing, and the very specific and important roles that they each play, into a common project.

In the end, I would say it’s less about the institutions and more about the people and finding the right people to work with. And it has been an amazing privilege to be able to scan across the entire landscape and find those people, to identify the group that wants to work together. That has been immensely sustaining throughout my career.

Getting the data right is a critical investment in the understanding that we’re building into machines, into AI, and our infrastructure.
Dr. Claire Melamed Vice President, AI and Digital Cooperation Strategy United Nations Foundation

What advice do you have for someone who wants to get involved in data and AI for social impact?

First of all: It’s great! Don’t be put off by all the doom and gloom: there’s no money, there are no jobs. There are still great, fantastic opportunities. I never would have envisioned the career I’ve had and all the twists and turns. AI in its current form obviously didn’t exist when I was in my 20s, but I benefited from being curious, open, and flexible—jumping on opportunities where they were presented to me.

I’ve never had a plan and, luckily, it’s worked quite well for me. Some people do have a plan, and that works quite well for them, but I often talk to young people who are worried that they don’t have a plan and think they should. I advise them to be open to opportunities when they come: to learn how to recognize them (which is not always easy!) and have the confidence to seize them.

What’s the next big thing you see for data and AI in the social sector?

I don’t think we’ve fully embedded this current phase of technology yet, so there’s a long way to go on adoption and diffusion.

We are still early for nonprofits to see what the current AI technologies can do for them. Today, it’s mostly in terms of embedding different AI tools in processes. That might sound boring, but it’s what will make the difference to the biggest number of people in the short term.

I don’t think we’re even at the end of the beginning in terms of what AI can do for the sector. There’s a level of AI adoption that is going to affect the whole structure of organizations, relationships in organizations, the kind of job profiles that organizations can offer, which then affects teams and how they work together.

We don’t really know what that’s going to look like, but I think it’s going to be big.

One thing I’m quite frustrated about in the AI debate, but I feel that people are starting to grasp, is the need for data governance. There’s a fairly general rule whenever anyone is talking about AI policy that at some point somebody will say, “and of course, data governance is really important.” And then there’s a pause, since everyone thinks that’s someone else’s problem.

But I’m hoping and starting to feel people are getting more serious about data governance at every level. You know, this is partly about the management of the data within any individual organization. People, of course, are more aware of that now, because there are more and more horror stories of where that’s gone wrong, and that’s concentrating minds on the issue.

The next step is to be a little bit more proactive than just reacting to a data problem. In the near term, I would hope to see a much more strategic, top-to-bottom approach to data governance. We need an approach that’s not defensive or reactive, but that’s actually grounded in the belief that if data is this powerful, if it is feeding the machine, we want to think about rights and accountability and oversight. We want to ensure that people have some sort of visibility into this data, which is shaping the algorithms that are shaping all kinds of everyday interactions and decisions.

A more strategic, values-focused approach to data governance is one thing that I’m starting to feel is possible in the near term, and certainly very, very necessary.

What’s your don’t-miss daily or weekly read?

I read a lot of books and alternate between the light and the serious in novels. I have a few classics, like Pride and Prejudice, that I regularly reread when I feel the need.

The most recent book I’ve read is This Is Where the Serpent Lives, which was fabulous. I spoke earlier about the ways data can show you lives that you didn’t understand. Good literature does that, too, and this is an absolutely fascinating book about Pakistan and the complicated dynamics across society. I also spend some of my time reading terrifying books on AI, like If Anyone Builds It, Everyone Dies. There is an emerging genre of these books, and we need to understand and take seriously these perspectives — even as we think about and plan for the good we hope data and AI can do.

About the Author

Perry Hewitt

Chief Strategy Officer

data.org

Chief Strategy Officer Perry Hewitt joined data.org in 2020 with deep experience in both the for-profit and nonprofit sectors. She oversees the global data.org brand and how it connects to partners and funders around the world.

Series