Discovery-Based Learning
Why learning by doing isn't just a philosophy. It's how the brain actually works.
What fifty years of research got right
Most corporate training is built on a transmission model: an expert stands at the front of the room and explains things clearly enough, and learning happens. It's efficient. It's scalable. And decades of research say it doesn't work.
In the 1990s, Richard Hodge founded The Real Learning Company on a different premise, one drawn from thinkers like Peter Senge, David Kolb, Jerome Bruner, John Dewey, and Lev Vygotsky, who had each arrived at the same conclusion from different directions: people learn by doing, not by being told. Knowledge is constructed through experience and social interaction, not transmitted through instruction.
This wasn't new science. Dewey wrote about learning by doing in the 1930s. Kolb formalized the experiential learning cycle in 1984. Vygotsky showed that social context is where higher-order thinking develops. But most of the training industry chose to ignore these findings because lecture-and-PowerPoint was cheaper to deliver.
Richard built programs that inverted the traditional model: participants started working immediately, with minimal instruction. Insights emerged through facilitated debriefs, not presentations. Small groups tackled real problems. The facilitator's job wasn't to deliver content. It was to create the conditions for discovery.
What brain science confirmed
Decades later, neuroscience caught up. Brain imaging technology made it possible to see why the learning theorists had been right all along, and the findings were striking.
Robert Bjork at UCLA discovered what he called "desirable difficulties," the counterintuitive finding that conditions making learning feel harder in the moment produce dramatically better long-term retention. Manu Kapur's research on "productive failure" extended this: learners who struggle with problems before receiving instruction outperform those who get instruction first. The struggle itself creates the neural structures that make the framework meaningful.
David Rock's work on the social brain revealed that the threat response (triggered by judgment, hierarchy, or fear of looking foolish) literally diverts oxygen from the prefrontal cortex, shutting down the brain's capacity for creative thinking and insight. Neuroscientist Naomi Eisenberger showed that social exclusion activates the same brain regions as physical pain. The implication: psychological safety isn't a nice-to-have in learning. It's a neurological prerequisite.
Meanwhile, research on surprise and learning showed that when expectations are violated, when something unexpected happens, the amygdala enhances attention and the brain encodes the experience more deeply. Every principle that Richard built on was now visible in brain imaging. The old theorists weren't guessing. They were describing the brain's actual architecture for learning.
And then AI showed up
Here's where Hi's story takes an unexpected turn. The same principles that describe how humans learn best turn out to mirror how artificial intelligence itself is built: neural networks, training through iteration, reinforcement learning, the value of diverse inputs. When Hi teaches people to work with AI through discovery, productive failure, and social learning, we're using a methodology that rhymes with how the technology itself was created.
That's not a coincidence. It reflects something fundamental about how intelligence works, whether biological or artificial.
In a Hi learning lab, participants use AI from minute one. They fumble with prompts before learning our PGA framework, and that productive failure creates the readiness for the framework to stick. They work on their actual business challenges, not simulations. They learn in small groups where honest conversation and cross-functional perspective create insights no individual could reach alone.
Jim Perry, who apprenticed with Richard in the 2000s before forging his own path at BTS (specializing in integrated learning journeys that wove together business acumen, people-leadership, and personal transformation) brought these threads together when AI arrived. Thirty years of learning science met the biggest workplace shift in a generation. And it turned out to be the exact right preparation.
See How It Works in Practice
What this looks like in practice
Do first, debrief second. Every Hi program inverts the traditional model. Participants start working immediately: with AI tools, with each other, on real problems. The big insights emerge through facilitated reflection, not front-loaded instruction.
The lab, not the lecture hall. Our learning environments are designed for experimentation, not presentation. Small groups. Hands-on activities. Real work. A space where it's safe to get things wrong, because that's where the deepest learning happens.
Social learning is structural. We deliberately vary between room-level, team, pair, and individual work because different kinds of thinking happen at each level. The cross-pollination between perspectives isn't a side benefit. It's the mechanism.
AHAs in the HAHAs. Fun and humor aren't decorative. They're structural. Laughter creates psychological safety, breaks down hierarchy, and opens people to new ideas. The best learning moments often come wrapped in surprise, play, or delight.
Reinforcement over time. A single experience, no matter how powerful, fades without reinforcement. Hi's Comets (short, daily reinforcement campaigns powered by the 1st90 platform) extend learning into the flow of work, building the neural pathways that turn insight into habit.
Want to see discovery-based learning in action?
Explore our solutions to see how these principles come to life, or get in touch to talk about your organization.