A Reflection on the Interplay Between Technology and Learning Theory

In a moment shaped by AI, I see many conversations focused on anticipating how the learning field will change—how it might reshape instructional design, learning experiences, and the role of educators. In times of uncertainty, I've often been able to learn a great deal by examining the past.

While learning theories are grounded in psychology and philosophy, the ways they are understood, applied, and extended have often been shaped by the technologies of their time. This is not to say technology determines theory, but rather that it often influences which theories gain traction, how learning is modeled, and how instruction is designed in practice.

So, this is my reflection on how specific technological shifts have influenced dominant learning paradigms and instructional design approaches over time. I hope it offers perspective on how we can respond to the changes AI brings—not just with urgency, but with purpose and care. And perhaps, with our priorities in the right place, we can use AI to benefit not only organizations, but people and society more broadly.

Behaviorism and the Logic of Early Instructional Systems

Behaviorism, with roots in the early 20th century, conceptualized learning as a measurable change in behavior. Learning occurred through stimulus-response conditioning and reinforcement. These principles aligned closely with the priorities of early instructional systems design, which emerged in industrial and military contexts where the goal was to train large numbers of people quickly, consistently, and with minimal variation.

Programmed instruction machines—like those developed by B.F. Skinner—embodied the behaviorist approach and matched the technical capabilities of the time. They delivered instruction in a linear, step-by-step format, provided immediate feedback, and emphasized mastery through repetition. Instructional design models during this period emphasized sequencing, clear objectives, and mastery through repetition—values well-suited to both mass education and early forms of instructional technology.

These principles continue to underpin many modern learning programs, especially in areas like compliance, technical skills, and procedural training—contexts where precision, repetition, and measurable outcomes remain essential.

Cognitivism and the Rise of the Computer as Metaphor

By the 1960s and 70s, as computer science advanced, so did models of human cognition. Learning was no longer seen just as behavior change, but as internal information processing—encoding, storing, and retrieving knowledge. The computer became both a metaphor and a practical tool. Cognitive load theory, schema theory, and information processing models aligned with how computers were understood to function: with working memory, long-term storage, and structured inputs and outputs.

Instructional design during this period became more systematic. Robert Gagné’s conditions of learning and Merrill’s component display theory focused on identifying and sequencing instructional events to support mental processing. These models laid the foundation for much of what instructional designers still use today—particularly when designing for complex knowledge acquisition, procedural tasks, and step-by-step skill development.

Cognitivist principles continue to guide modern design practices, especially in contexts like onboarding, technical training, and performance support. Techniques such as chunking information, managing cognitive load, and supporting transfer through practice and feedback all trace back to this era’s emphasis on how learners process and organize knowledge internally.

Constructivism in Practice: Empowered by Digital Tools

Constructivism, which emphasizes active meaning-making and learning through experience, long predates the digital era. Its roots are philosophical and social—grounded in theorists like Piaget and Vygotsky, who emphasized that learners build knowledge through exploration, dialogue, and reflection.

While the theory itself is not technology-driven, the rise of multimedia, hypertext, and interactive digital tools created new opportunities to apply constructivist principles at scale. Learners could now explore, manipulate, and co-create meaning in environments that weren’t possible through print-based or linear instruction.

Today, constructivist approaches remain influential in instructional design, particularly in contexts where the goal is to build critical thinking, problem-solving, or collaboration skills. Scenario-based learning, project-based learning, and authentic assessments all draw from constructivist thinking. Digital tools—such as branching simulations, collaborative platforms, and sandbox environments—continue to expand the ways learners can engage with content and construct understanding.

Where AI Is Taking Us—and What Past Shifts Help Us Understand

Each major learning theory emerged in response to what people needed to do. Behaviorism supported task execution, cognitivism emphasized mental organization, and constructivism prepared learners for meaning-making in complex environments. As technology advanced, expectations moved up the value chain—and instructional design evolved accordingly.

AI continues that shift. By automating tasks such as content summarization, data classification, and even insight generation, AI changes what people are expected to contribute. The result is a heightened emphasis on higher-order capabilities—judgment, ethical reasoning, and contextual thinking—which are becoming central to what learning must now support.

What People Need to Learn Is Changing

Instructional goals must now reflect a new kind of human-AI partnership. When machines can retrieve, synthesize, and generate information, people are expected to go further: to frame problems, interpret outputs, evaluate quality, and make decisions under uncertainty.

This redefines what competence looks like in many domains. Designers must shift focus from knowledge recall and procedure-following to developing skills in oversight, reasoning, and responsible application. Instructional models will need to make space for ambiguity, ethical complexity, and interpretive thinking—areas that cannot be easily automated.

How We Design and Deliver Learning Is Expanding

At the same time, AI is expanding what’s possible in learning design both in how content is personalized and how social interaction is simulated.

Scalable Personalization of Cognitive Design: Adaptive systems now allow content, pacing, and feedback to adjust in real time. This makes long-standing principles, like scaffolded feedback, spaced retrieval, and differentiated progression, more feasible to implement at scale. AI strengthens our ability to apply behaviorist and cognitivist strategies more precisely and responsively, improving alignment between learner needs and instructional delivery.

New Affordances for Social and Situated Learning: AI tools are also beginning to approximate elements of social interaction. Conversational agents, collaborative bots, and simulated peers can model dialogic learning and peer engagement. While not replacements for human interaction, they allow instructional designers to incorporate principles from situated learning and sociocultural theory, enabling forms of participation, feedback, and meaning-making that previously required live group settings.

Final Thoughts

Looking at the history of learning theory through the lens of technology reveals a consistent pattern: as the tools and demands of work evolve, so too does the focus of instructional design. Theories don’t vanish; they adapt. Behaviorism didn’t disappear when cognition became central, and constructivism didn’t replace memory models—it simply offered a different lens for more complex tasks.

AI fits within this trajectory. It shifts the ground beneath how existing ones are used. It prompts us to teach different things—judgment, reasoning, interpretation—and gives us new tools to design learning that is more adaptive, responsive, and socially nuanced.