DataKind’s Founder & Executive Director, Jake Porway, and Nick Hamlin, Data Scientist at GlobalGiving recently caught-up in DataKind’s offices for a chat. Here’s some of what they spoke about…
A little about Nick…
- Data scientist, worm farmer and lives to make social good data accessible, understandable, and actionable for everyone.
- Masters of Information and Data Science from UC Berkeley; degrees in Mechanical Engineering and History from the University of Rochester.
- Developed Pando, Root Change’s platform for exploring networks in the social sector.
- Outside the office, Nick’s a folk music street performer, amateur haiku poet, and, did we mention, a worm farmer?
Jake: What inspires you to do the work you do?
Nick: I’m fortunate enough to get to support some of the most inspiring change-makers on the planet through my work at GlobalGiving, so it’s easy to get excited about coming in to work everyday. There’s also so much new technology available in the world of data science that carries vast potential for the social sector, but many nonprofit organizations have a hard time navigating the landscape because of how fast it’s growing and changing. This leads to lots of exciting opportunities for data scientists to provide thoughtful guidance and expertise to the social sector to help bridge this gap. It’s a unique skill set though, requiring tech fluency and comfort with listening to and learning from real people. Fortunately, organizations like DataKind do a phenomenal job finding folks to fill these roles!
Jake: Thanks Nick, that’s a big part of our mission. I’m curious, can you describe GlobalGiving’s mission and how AI and machine learning help you meet the organization’s goals?
Nick: GlobalGiving is here to transform aid and accelerate community-led change. That’s a tall order, especially given the diversity of programs and geographies that our thousands of partner organizations work in and the relatively small size of our team (roughly 50 people). As a result, we love finding opportunities where we can use technology like AI and ML to supercharge the work that our staff can do. Informally, we’ve referred to this as “making people’s jobs harder”, which sounds counterintuitive. What we’re actually doing is taking a lot of the tedious and rote work that can be solved with an algorithmic approach off of the plates of our staff so that they can focus on the grey areas – those tough questions that come up all the time in international development work that require a uniquely human mind to solve.
Jake: Speaking of hard jobs, disaster recovery is a significant piece of GlobalGiving’s work. Can you give any current examples this work?
.png)
Nick: Yes! The Disaster Recovery Network at GlobalGiving is one of our busiest areas of work. We focus on supporting disaster recovery work led by members of the affected communities, since these are the change-makers who have the nuanced and in-depth understanding of on-the-ground needs and will be there supporting the long-term recovery for years after the headlines have died down.
When an event like Hurricane Florence happens, every part of GlobalGiving springs into action, from members of our program team coordinating with the local nonprofits who are responding, to our communications team fielding questions from journalists covering the storm, to our operations team who typically have the first disbursements of funds out the door within a week after the event. There’s lots more information available on our Disaster Recovery Network page as well.
Jake: That’s obviously very topical and important work. So if you had the power of AI-for-Good in your total command, what problem would you solve first?
Nick: I think it’d be extremely transformative for the social sector if we had a rigorous and scalable way of defining impact. There’s been tons of work done in this field, including at GlobalGiving. The tension between AI’s need for clear outcome metrics and the social sectors nuances is real. When I tell people about the work I do in social sector data science, I frequently get responses like, “oh, so you’re trying to ‘Moneyball’ the social good world”? I usually push back a bit in these situations.
The key difference is that everyone has a clear and agreed upon understanding of what it means to win a baseball game, so any data-driven solution in that world already has agreement on the problem being solved. In contrast, we don’t have anywhere near the same clarity on what it means to create “good” social change, and it may not even be possible to arrive at such a “grand unified theory of good”. As a result, it’s critical for data scientists in the social sector to take the time to make sure they’re asking the right questions before building the algorithms to try to find the right answers.
