Solving Public Problems with Artificial Intelligence
How to take advantage of technology, data, and the collective wisdom in our communities (and now AI?) to design powerful solutions to contemporary problems
Open Minds and Boundless Energy
Earlier this Fall, young people turned out to vote in record numbers in elections in New York City, Virginia, and New Jersey. From gun rights activism after the Parkland mass shooting to the GenZ revolution in Nepal, young people are engaging in their communities. As Harvard researchers Felton (“Tony”) Earls and Mary Carlson point out in Voice, Choice and Action, “as the world’s population ages, societies will depend ever more on the open minds and boundless energy of children.”
In this time of great challenges, our democracies urgently need to produce citizens who can move from demanding change to making it. But the skills for doing so are not innate; they are learned.
Students who attend schools that provide civic skills training are more likely to be civically and politically engaged outside of school, researchers have found. However, too often, our schools and universities are failing to provide training in how to effect social change. Although problem-solving has been identified as one of the most essential skills a graduate needs in the 21st century, educational institutions rarely teach it.
Even where schools offer entrepreneurship or design, helping students imagine good ideas, they are often not equipping them at scale with the skills to implement those ideas in the real world.
Even where schools offer entrepreneurship or design, helping students imagine good ideas, they are often not equipping them at scale with the skills to implement those ideas in the real world.
At this time of rising political violence and frustration, universities need to give students a constructive outlet for making change happen, and artificial intelligence offers a new opportunity to do so at scale and with greater impact.
The First Solving Public Problems Course
When Covid-19 exploded in early 2020, NYU wanted to offer cooped up students ways to make up for the disheartening experience of missing out on campus life. Thus, I created Solving Public Problems to provide “a systematic introduction to the problem-solving skills students will need to take a mission-driven project from idea to implementation.”
The seven-week online course is freely available online here at solvingpublicproblems.org. Both are based on the eponymous book.
After several years and thousands of learners in a hundred countries, the course has proven that people can be taught the skills of tackling wicked problems.
The course covers a dozen problem-solving competencies. Course lectures are accompanied by interviews with 30 leading global change-makers from the head of the United Nations Development Programme to the co-founder of the Movement for Black Lives, discussing their careers in mission-driven work.
Of the 200+ undergraduate students from 48 countries who signed up for the first iteration of the course, over two-thirds said they were unfamiliar with the skills of problem solving. Half said they had not or were unsure if they had learned quantitative methods, and another 58% any qualitative skills while in university.
When asked, “Why are you taking the course?” one student replied, “Every single course that I've had has always been based on theory; there hasn't been a single course that actually deals with practical problems that arise. Reading about Fanon or Gandhi was wonderful, but how do we [if possible] apply their theories into real life? Such questions have always bugged me, and having a course that directly deals with 'public' life – a life beyond books – is what inspired me. I look forward to learning how to do real stuff, in real-world, to actually make an impact.”
After learning how to break intractable issues into manageable problems in consultation with those affected by the problem; how to use data and collaboration to understand those problems as well as techniques for conducting rapid evidence reviews to scavenge for solutions that have worked elsewhere as well as how to write a persuasive letter, craft a professional memo and make a killer deck, 98% said they learned more about how to take a mission-driven project from idea to implementation.
“Every single course that I've had has always been based on theory; there hasn't been a single course that actually deals with practical problems that arise. Reading about Fanon or Gandhi was wonderful, but how do we [if possible] apply their theories into real life? Such questions have always bugged me, and having a course that directly deals with 'public' life – a life beyond books – is what inspired me. I look forward to learning how to do real stuff, in real-world, to actually make an impact.”
After several years and thousands of learners in 100 countries, the course has proven that people can be taught the skills to tackle wicked problems. But what used to take weeks of research, analysis, and outreach can now be accelerated with artificial intelligence. The same methods we practiced manually—breaking problems into parts, uncovering their causes, gathering data, and talking to those affected—can now be jump-started and scaled with AI.
How Artificial Intelligence Can Help
That’s why I am now remaking the Solving Public Problems course for the AI era.
When learners begin the problem-solving process, their first and most challenging task is to define the problem clearly. That means understanding why it’s happening and for whom. In the past, uncovering the root causes of an issue required long hours of searching for studies, comparing data, and talking to experts and the affected public. Today, AI can act as a catalyst in that process, helping us explore potential causes, test hypotheses, and see patterns in the data before we head into the field.
Take the example of school truancy. I was working with leaders in a state (no, not New Jersey) who, having recently learned about behavioral insights, were eager to send text reminders that school attendance is required. When I asked why they thought kids were staying home, they responded: “Their parents want to get discounts at Disneyland by visiting in the off-season.”
Their solution—sending text reminders—made sense only if that assumption were valid. With AI tools, we can now test that hypothesis quickly: What are the actual reasons behind absenteeism? AI can surface relevant research and data pointing to rural geography, poverty, or transportation barriers. Instead of substituting for human judgment, AI helps refine our understanding through rapid research, enabling interventions to be more evidence-based, targeted, and effective.
But data alone is not enough. We also need to understand people’s lived experiences. In projects like Code for America’s redesign of California’s food assistance program, listening to applicants revealed that the forms themselves were the main barrier. Now, with AI tools that can summarize thousands of interviews or community meetings, it’s possible to scale this kind of listening.
Remaking the Course
Used wisely, AI could allow more people to practice the democratic arts of problem-solving at a new scale. That’s why I am now remaking the Solving Public Problems course for the AI era. The core skills remain the same, but new technologies make it easier to teach and practice them. With AI, we can help learners find relevant data faster, conduct rapid evidence reviews, summarize community feedback, and visualize ideas more vividly. The goal is not to shortcut the human work of engagement, but to support it—to use AI to listen better, to include more voices, and to reach further into communities that have too often been left out.
As I revise the course, I’m eager for feedback:
- What skills do you believe today’s students—and tomorrow’s public problem solvers—most need to learn to become effective agents of social change?
- How should we think about the opportunities and risks of introducing AI into civic learning?
- How can we ensure that technology strengthens, rather than substitutes for, our capacity to connect, deliberate, and act together?
If, as Bertrand Russell wrote, “power is the ability to produce intended effects,” then our task as educators is to make people more powerful in their own lives and communities. Teaching problem solving—now with the thoughtful use of AI—can help us do just that.