Problem-Based Learning and Future-Ready Skills

Even if the most dire predictions about AI fail to materialize, there’s no doubt that its arrival raises big questions about the purpose of K–12 education. Most existentially: What skills do students need for an uncertain future?

My experience and research suggest that the answer isn’t much different than it was: Students need the triple skill set of communication, collaboration, and critical thinking. And more than ever, they need the opportunity to grapple with real-world problems. The best structure I’ve seen for teaching, developing, and assessing the skills that students are going to need in an AI world is problem-based learning.

Real-World Problems

We know that students engaged in problem-based learning perform better academically, but we also know that they’re more career-ready, better collaborators, and more equipped to confront the challenges of tomorrow. AI’s promise is efficiency, optimization, a frictionless life. But that’s not a lived experience. “In recent years, society has been conditioned to see friction not as a teacher but as a flaw—something to be optimized away in the name of efficiency,” technology executive Raffi Krikorian writes in The Atlantic. “But friction is where discernment lives. It’s where thinking starts.” Well-planned and executed problem-based learning has the potential to be this teacher, to let students experience the friction required for the kind of learning that prepares them for the future.

Authentic, real-world problems matter. But so does how we approach them. I’ve worked with problem-based learning—both in my own classroom and as part of two programs based on it—and the most successful projects are the ones in which the problems are real and students actually have to try to solve them. Not theoretically and not in a slide show presented to their teachers. ChatGPT can do that if that’s all we want. What students need to be able to do is navigate the hundreds of tiny obstacles that exist between a problem and its solution.

AI can’t understand the application of school policy, or culture or personality clashes. AI can’t know who holds the knowledge that students need or the state of equipment they require. Students must have the space and time to actually try to implement their ideas, fail, and try again. And for this they need creativity.

Critical Thinking

Problem-based learning offers a space for ideation, iteration, and exploration of perspectives. This is the type of creativity that students need to practice, that they need to become comfortable with to tackle actual issues. But creativity doesn’t happen in a void. It requires context and information and the ability to make sense of that information. In other words, students also need to be able to think critically about their topic.

The best explanation of critical thinking I’ve seen is from Kaveh Akbar’s novel, Martyr! “Well, of course I changed my mind, I was presented with new information,” a character says, “that’s the definition of critical thinking.” AI is going to present new information and will (hopefully) get more and more accurate with this information. But are we developing students who are nimble enough to shift their thinking as they learn more, as they confront obstacles, as the context that they’re working with changes? If not, we need to rethink how we’re structuring their learning.

Collaboration

Another key component of problem-based learning that I see as essential for existing in an AI future is collaboration. In my years in these programs, this is the skill in which I’ve seen the most growth and the most interdisciplinary transfer. We can’t just ask students to do group work and expect good things to happen anymore than we can give them a math problem and simply say, “Solve it.” Decision-making strategies, conflict resolution, team-skill analysis: These all need to be explicitly taught.

With teaching and practice, students learn each other’s strengths, learn to trust each other, learn to work with people who are different from themselves. AI can be a thought partner (despite its current sycophantic tendencies), and it can build and guide a process. But it can’t achieve the kind of collaboration that happens when there’s tension and healthy conflict and competing perspectives. It can’t know that one approach is better than another because of the dynamics and passions of that exact combination of students.

Communication

The final component of problem-based learning that I see as being necessary to developing future-ready students is the ability to communicate their ideas to an audience of experts. Often, this experience is mimicked with peers and teachers acting as experts, and that can be OK. But presenting to real experts requires a richer understanding of the topic, of the vocabulary, of the challenges in that field. It forces a more refined and specific message.

And most important, it forces students to grapple with real feedback. Instead of the final component of a project, it becomes an opportunity to deepen their understanding and continue to solve the problem. In the most enriching projects I’ve seen, students “end” with a whole new set of questions that they couldn’t possibly have predicted at the outset.

Problem-based learning is not a magical cure-all. But it is the most authentic approach to preparing students for an uncertain future, especially when projects are designed with the space for students to hit real-life obstacles and work through them. The threat of AI-completed tasks and assessments is real, but I think a much bigger threat is allowing students to enter a future that no longer values the skills acquired in a traditional classroom. We can’t put the AI genie back in the bottle and pretend that this is not having a dramatic and immediate impact on our world. So we need to develop the kinds of units, courses, and programs that give students the best chance of navigating the future ahead.