Why do so many people give up when learning something new, even when they're perfectly capable? It's not a lack of intelligence, but a fundamental bottleneck in how the brain processes information - a cognitive traffic jam that occurs when we try to learn in ways that don't align with how our brains are designed. Two scientists set out to understand this frustrating experience, and in doing so, built an online course that has reached millions.
Barbara Oakley knows this feeling intimately. She flunked her way through early math classes. She describes herself, with some understatement, as someone who did not naturally connect with quantitative subjects. And yet she became a professor of engineering, a published researcher in neuroscience and social behavior, and the co-creator of one of the most enrolled online courses in history. The distance between those two realities the struggling student and the award-winning professor is not a story of redemption through willpower. It is, according to Oakley and her collaborator Terry Sejnowski, a story about how the brain actually learns, and how understanding that process can change who learns what, and when.
"We created one of the world's most popular MOOCs," Oakley and Sejnowski wrote in a 2019 paper published in npj Science of Learning, a peer-reviewed journal from Springer Nature. That paper, titled "What we learned from creating one of the world's most popular MOOCs," documented not just the scale of their course but the scientific principles behind its design. Learning How to Learn: Powerful mental tools to help you master tough subjects, offered through Coursera, had already reached millions of learners across borders, languages, and age groups. What made it different, they argued, was that it was built on neuroscience not just teaching philosophy or motivational rhetoric, but the actual mechanics of how neural circuits form, strengthen, and transmit knowledge.
The collaboration that produced this course is, itself, a story about unexpected connections. Oakley, the engineering professor with the unconventional path, and Sejnowski, the computational neuroscientist who helped pioneer neural network research at the Salk Institute, come from different intellectual traditions. Oakley's work focuses on the intersection of neuroscience and social behavior. Sejnowski's research lives in computational modeling of brain function. Together, they built something that neither could have built alone: a course that takes the laboratory and translates it into a living curriculum.
The Two Minds: Focused and Diffuse
At the heart of Learning How to Learn is a concept that sounds almost too simple to be revolutionary: the brain has two distinct learning modes, and understanding when to switch between them is the key to mastering difficult material. Oakley and Sejnowski call these the focused mode and the diffuse mode. The focused mode is what happens when you concentrate directly on a problem tracing equations, parsing grammar, memorizing vocabulary. The diffuse mode is what happens when you step back. When you go for a walk. When you let the mind wander. Research suggests that the diffuse mode is not laziness. It is consolidation. The brain, given space, quietly reorganizes what it has absorbed, forming connections that focused attention alone cannot forge.
This is not vague educational philosophy. It is grounded in how neural circuits actually work. In a February 2025 episode of the Huberman Lab Podcast, hosted by Andrew Huberman, Terry Sejnowski explained the neuroscience behind this dual-process model with unusual clarity for a general audience. The conversation, which explored "The Science of Learning A Neuroscience and AI Perspective," touched on what Sejnowski calls the algorithmic level of brain function the layer between physical structure and observable behavior where neural circuits process information the way a recipe transforms raw ingredients into a finished dish.
"The brain uses algorithms, much like a recipe, to achieve specific outcomes," Sejnowski explained in the Huberman Lab discussion. "These algorithms are crucial for understanding how we learn and adapt." This framing the brain as a system that runs procedures, not just stores facts is central to both his research and to the course he co-designed with Oakley. Learning is not absorption. It is execution. And like any execution, it can be understood, practiced, and improved.
The Basal Ganglia and the Chemistry of Repetition
Sejnowski's own research career offers a window into how these neural algorithms were discovered and named. Trained in physics at Case Western Reserve University and Princeton, where he earned his PhD in 1978, Sejnowski pivoted toward computational neuroscience the use of mathematical models to understand how the brain processes information. His doctoral advisor was John Hopfield, whose Hopfield networks became foundational work in neural computation. Sejnowski's own contributions include the development of Independent Component Analysis and early work on the Boltzmann machine, both of which helped researchers understand how neural networks could learn patterns without explicit programming.
But for the purposes of Learning How to Learn, one of Sejnowski's most important contributions is his research on the basal ganglia a set of brain structures tucked below the cortex that play a central role in learning action sequences and refining movements through repetition. In the Huberman Lab discussion, Sejnowski described how the basal ganglia acts as a kind of training coach for neural circuits. When you practice something whether it is a tennis serve or a calculus technique the basal ganglia helps the brain automate the process, shifting control from conscious, effortful cortical processing to smoother, more efficient subcortical routines.
This mechanism is why deliberate practice matters. It is why cramming, despite feeling productive, rarely produces durable knowledge. And it is why spacing out learning sessions allowing time for the diffuse mode to consolidate what the focused mode has absorbed produces better long-term results than massed practice. The basal ganglia does not work on demand. It works on schedule, consolidating information during rest, sleep, and the mental wandering that most people mistake for wasting time.
Barbara Oakley's Unconventional Path to Learning Science
The personal history behind Learning How to Learn is not incidental. It is, in many ways, the pedagogical engine of the course. Oakley did not arrive at neuroscience through a straight academic path. She rose through the ranks of the U.S. Army from Private to Captain, earning recognition as a Distinguished Military Scholar along the way. She worked as a communications expert at the South Pole Station in Antarctica. She served as a Russian translator aboard Soviet trawlers on the Bering Sea. These are not the credentials of a typical professor of engineering, and they are not presented as biographical curiosities they are, in Oakley's telling, direct laboratories for understanding how humans adapt, learn, and perform under extreme conditions.
"Her work focuses on the complex relationship between neuroscience and social behavior," according to her instructor profile on Coursera. That focus on the intersection of brain science and human behavior shapes how she teaches. Oakley's research has been described as "revolutionary" in the Wall Street Journal. She has published in the Proceedings of the National Academy of Sciences, the Wall Street Journal, and The New York Times. She holds the McGraw Prize often referred to colloquially as the "Nobel Prize for Education" and is an elected Fellow of both the Institute of Electrical and Electronic Engineers and the American Institute for Medical and Biological Engineering. These are not soft credentials. They are the institutional recognition of someone whose research has measurably advanced the science of learning.
Her books extend this work into accessible territory. A Mind for Numbers: How to Excel at Math and Science (Even If You Flunked Algebra) applies the neuroscience behind Learning How to Learn specifically to quantitative subjects. Mindshift: Break Through Obstacles to Learning and Discover Your Hidden Potential expands the framework to career changes and lifelong learning. Uncommon Sense Teaching, published by Penguin Random House in 2021, brings the principles into classroom and online teaching contexts. Together, these books and the course form a connected body of work each component reinforcing the others, each translating the same underlying science into different applications.
The MOOC Experiment: What One of the World's Most Popular Courses Taught Its Creators
The 2019 npj Science of Learning paper that Oakley and Sejnowski co-authored is notable not just for documenting the course's success but for reflecting honestly on what they learned from it. A MOOC Massive Open Online Course is, by design, a democratizing technology. It removes geographic barriers, financial barriers, and institutional barriers. Anyone with an internet connection can enroll in a course taught by world-class instructors. But scale changes the pedagogical challenge. A lecture designed for a classroom of forty students behaves differently when delivered to four hundred thousand. Interaction patterns shift. Feedback loops become noisy. The assumptions instructors make about learner context about time, about prior knowledge, about motivation stop holding.
Oakley and Sejnowski write in their paper that the process of building Learning How to Learn for such a vast and varied audience forced them to distill learning science to its most essential elements. The neuroscience could not be simplified into vague inspiration. It had to be specific enough to be actionable, clear enough to be remembered, and broad enough to apply across disciplines ranging from music to medicine. The focused and diffuse mode framework emerged, in part, from this pressure it is a concept that is neuroscientifically grounded, immediately graspable, and universally applicable.
The course is not, despite its popularity, a replacement for formal education or personal mentorship. But it is, its creators argue, a foundation a set of mental tools that changes how learners approach any subsequent learning challenge. Once you understand that your brain is designed to consolidate information during rest, you stop feeling guilty about taking walks during study sessions. Once you understand that repetition and spaced practice build the basal ganglia routines that underlie expertise, you stop expecting fluency after a single pass through difficult material. The course does not teach specific content. It teaches the operating system under which all content is learned.
Sejnowski's Algorithmic Brain and the AI Connection
The Huberman Lab conversation between Sejnowski and Andrew Huberman covered territory that extends beyond traditional education science into the relationship between neuroscience and artificial intelligence a connection that is not incidental to Sejnowski's research career. Sejnowski's work on neural networks, dating back to the 1980s, helped establish the theoretical foundations for the kind of machine learning systems that now power everything from language translation to image recognition. His perspective on learning is, by design, interdisciplinary drawing on computational models that were originally inspired by biological neural circuits and now inform those same biological models in return.
In the Huberman Lab discussion, Sejnowski described how the algorithmic level of brain function sitting between the implementation level (the physical mechanisms) and the behavioral level (what organisms do) is where the most productive research questions live. This framing, which he has advocated for in the scientific literature, is essentially the same framing that powers modern AI: understanding not just what a system does but what computational process it runs to produce that behavior. When Sejnowski describes the basal ganglia as a system for learning sequences and refining them through repetition, he is describing both a biological mechanism and a principle that AI researchers have independently discovered: that performance improves through iterative practice guided by reward signals.
This convergence is not metaphorical. It is structural. The same mathematical frameworks that describe reinforcement learning in artificial systems describe what the basal ganglia does in biological ones. Sejnowski, more than most neuroscientists, has spent his career tracing these connections and more than most AI researchers, he brings the biological depth to bear on questions of learning that purely computational approaches might miss.
Why This Matters for KnowledgePosts Readers
For readers researching practitioners, frameworks, books, and ideas the core audience of KnowledgePosts Oakley and Sejnowski's work represents something relatively rare in the learning resources space: a framework that is both rigorously grounded in peer-reviewed neuroscience and genuinely accessible to non-specialists. The course, the books, and the underlying research form a connected architecture. You can start with the Coursera course and move into Oakley's books. You can read the npj Science of Learning paper for the pedagogical data. You can listen to the Huberman Lab episode for a conversation that connects learning science to AI and cognitive enhancement.
What makes this especially valuable for KnowledgePosts readers is the specificity. Learning How to Learn is not a vague self-improvement course. It does not promise transformation through positive thinking or habit stacking alone. It offers named, explained, neuroscience-backed techniques: the focused and diffuse mode framework, the role of the basal ganglia in skill automation, the importance of spaced repetition and interleaved practice. These are not opinions. They are mechanisms described in the scientific literature, tested in the course's massive enrollment data, and translated into a curriculum that has reached millions of learners across languages and borders.
For anyone building learning resources, designing training programs, or studying how expertise develops, the Oakley-Sejnowski collaboration is a case study in what happens when laboratory science meets educational practice without compromising either.
Where the Work Stands in 2026
As of August 2026, Learning How to Learn remains available on Coursera. Oakley has continued to publish and expand the framework, with additional courses covering critical thinking, logic, and teaching methodology. Sejnowski continues his research at the Salk Institute and UC San Diego, where he serves as co-director of the Institute for Neural Computation. In 2025, he was elected to the American Philosophical Society a recognition of a career that has spanned neural network theory, computational neuroscience, and the science of learning at scale.
The course's enrollment numbers, which already placed it among the world's most popular MOOCs, have continued to grow as online learning has become a standard part of how people develop new skills. The framework has been adapted for youth learners, for professional development, for language learning evidence that the underlying principles transfer across contexts in the way Oakley and Sejnowski intended.
Key Elements of the Oakley-Sejnowski Learning Framework
| Concept | Description | Source |
|---|---|---|
| Focused Mode | Direct, concentrated attention on a specific problem or concept what most people think of as "studying" | Learning How to Learn (Coursera) |
| Diffuse Mode | Relaxed, unfocused mental state that allows the brain to consolidate information and form unexpected connections | Learning How to Learn (Coursera) |
| Basal Ganglia Role | Brain region that automates learned sequences through repetition, shifting control from conscious effort to efficient routine | Huberman Lab discussion (2025) |
| Algorithmic Level | The layer of brain function between physical implementation and behavioral output where neural circuits process information like computational procedures | Huberman Lab discussion (2025) |
| Spaced Repetition | Practicing material across distributed time intervals more than massed cramming leverages diffuse mode consolidation | Learning How to Learn (Coursera) |
| Interleaving | Mixing different types of problems or skills within a single study session more than blocking them improves long-term retention | A Mind for Numbers (Penguin, 2014) |
Where to Read Further
For readers who want to go directly to the source material, the following resources offer the most direct access to Oakley and Sejnowski's work and research:
- npj Science of Learning: "What we learned from creating one of the world's most popular MOOCs" the peer-reviewed paper by Oakley and Sejnowski documenting the course's design, scale, and pedagogical lessons
- Dr. Barbara Oakley's Coursera instructor page listings for Learning How to Learn and her full suite of related courses including Mindshift, Uncommon Sense Teaching, and Critical Thinking
- Huberman Lab: The Science of Learning - A Neuroscience and AI Perspective a conversation between Terry Sejnowski and Andrew Huberman exploring the algorithmic level of brain function, the basal ganglia, and AI-learning connections
- Terry Sejnowski - Wikipedia biographical and research overview including his work on neural networks, the Boltzmann machine, and his current positions at the Salk Institute and UC San Diego
The books A Mind for Numbers, Mindshift, and Uncommon Sense Teaching, all published by Penguin offer extended, book-length development of the course themes and are widely available through public libraries and booksellers.