Notice
Find the oversimplified idea: completion as proof, content as coverage, polish as quality, or AI output without judgment.
Perspectives
This is where I work through the questions that keep coming up in learning design. Why do we jump to courses? What should AI actually help with? How do we know our work made a difference? Come for a point of view, stay for the practical ideas you can take back to work.
Find your argument · 20 seconds
Editorial position
Every Perspective starts with a problem I recognize from the work, challenges the assumption underneath it, and turns that thinking into a move you can actually use.
Find the oversimplified idea: completion as proof, content as coverage, polish as quality, or AI output without judgment.
Question the default shape. Ask whether the real answer is a course, a system, a tool, a workflow, or support in the moment.
Turn the argument into a method a working learning designer can actually pick up and use.
Argument library
AI is not eliminating the need for instructional designers. It is exposing which parts of the job were production tasks all along.
Read the essayUsing AI to accelerate the same course-first process gives you more output, not necessarily better performance.
Read the essayThe LMS still matters for assignment, records, and compliance. It fails when we ask it to become the front door to every moment of support.
Read the essayTraining builds capability before the moment of performance. Performance support helps someone act during it. The right answer depends on the work.
Read the essayUseful L&D agents should reduce friction around decisions, evidence, and workflow support instead of becoming another chatbot nobody needs.
Read the essayPrompt technique changes quickly. Judgment is what helps a designer decide what to ask, what to trust, and what to build.
Read the essayPrompt engineering can improve today's output. It is too fragile and tool-dependent to become the foundation of an L&D strategy.
Read the essayThe learning systems designer connects courses, support, tools, managers, data, and workflow into an environment where performance can improve.
Read the essayWhen information is instantly available, learning should focus more on mental models, judgment, practice, and knowing when the answer cannot be trusted.
Read the essayMany learning strategies keep adding assets without deciding what deserves maintenance, retirement, connection, or a clear moment of use.
Read the essayAs AI handles more routine production, the value of empathy, facilitation, judgment, influence, and sensemaking becomes easier to see.
Read the essayEmbedded guidance, AI assistance, and connected tools can bring support into the workflow, but only when the experience is designed around the task.
Read the essayCourses still matter, but they should be chosen because sustained practice and structured learning are necessary, not because L&D needs an artifact.
Read the essayWhen tools, incentives, or the environment make good performance difficult, another course only teaches people how the system wishes work happened.
Read the essayA completion rate tells you that people reached the end. It does not tell you whether they can perform, whether behavior changed, or whether the business problem moved.
Read the essayAI accelerates outlines, scenarios, quizzes, visuals, and production. It can also turn a weak assumption into a finished course before anyone challenges it.
Read the essayWhen support lives far away from the work, people stop hunting for it. Design resources around the moment of need, not the organization chart.
Read the essayA beautiful job aid is useless when workers cannot locate it, trust it, or use it quickly inside the real workflow.
Read the essayWhen experts review a polished course, feedback drifts toward wording, colors, and preference. Give them the decisions only they can validate.
Read the essayLearning teams often treat release day as completion. Product-minded learning uses launch to begin observing, improving, and supporting the experience.
Read the essayLearning becomes bloated when every stakeholder request is equally important. Design requires hierarchy, tradeoffs, and the courage to cut.
Read the essayLearners bring anxiety, past experiences, interruptions, and uncertainty. Use a five-question learner-context review to reduce avoidable friction without lowering the standard.
Read the essayA finished sample shows what you built. A chain of evidence shows how you investigated the problem, chose a response, tested it, and understood its limits.
Read the essayA story and three choices can still test answer recognition. Use a decision audit to examine the cues, information, tradeoffs, and feedback your scenario actually provides.
Read the essayOpen-ended requests for feedback invite conflicting edits. Give each review a decision, an evidence standard, and an owner who can resolve disagreement.
Read the essayPrompting is not just a technical trick. For learning designers it is becoming a way to clarify intent, constraints, learner needs, and judgment before generating output.
Read the essayA content dump is not a learning plan. Use five buckets to decide what should be taught, practiced, supported, referenced, or cut before you start building.
Read the essayVisual polish can attract attention, but it cannot prove that a learning solution addresses the right problem. Use the five-question beauty contest test to examine what the work helps people do, what is blocking performance, and what evidence would show improvement.
Read the essayAI can generate nearly every visible part of a course. The harder and more valuable work is deciding whether a course is the right response to the real performance problem.
Read the essayAdding another course without a plan for ownership, review, consolidation, and retirement makes the learning library harder to navigate, maintain, and trust.
Read the essayCourse length is usually the wrong starting point. Before deciding five minutes or sixty, clarify what people need to do differently when the experience is over.
Read the essayLearning works better when it is treated less like a one-time deliverable and more like a living product experience built around usability, support, maintenance, and real work.
Read the essayLaunch is not the finish line. Learning designers should borrow the marketer's instinct for journeys, repeated touchpoints, reinforcement, and what brings people back after the first interaction.
Read the essayLearning design has to account for forgetting. If people need to remember, decide, or act later, the experience needs retrieval, reinforcement, and support beyond first exposure.
Read the essayCorrect and incorrect are not enough when the goal is better judgment. Feedback should help learners understand what happened, what they missed, and what to try next.
Read the essayA course can be part of the system, but it should not be mistaken for the whole strategy. Better design considers practice, support, manager behavior, tools, feedback, and evidence.
Read the essayPedagogy and andragogy are not interchangeable. Understanding the distinction is foundational, but modern workplace learning must ultimately be designed for performance, not just instruction.
Read the essayEvaluation fails when treated like a post-launch reporting ritual. Better evidence starts while the problem, behavior, and system are still being defined.
Read the essayPerspective lanes
Learning is not an event. It is a system of practice, feedback, tools, support, and evidence.
Measurement works better when it starts before launch, not after the course is already built.
Retention needs retrieval, reinforcement, transfer, and support at the moment of need.
Feedback should help learners reason, recover, and improve, not just reveal whether they were right.
AI can accelerate output, but judgment determines whether the learning actually works.
Learning should be designed with usability, iteration, evidence, and real workflow behavior in mind.
From argument to action
Perspectives name the argument. The Lab helps you pressure-test the thinking, design the response, and put it to work.
Learning, Rewired · the cadence
Learning, Rewired is the LinkedIn newsletter version of this thinking. A newsletter issue can grow into a Perspective, and a Perspective can become a newsletter issue: the same arguments, moving back and forth between the two.
Learning Rewired Lab
Use the thinking to challenge the default, make a stronger decision, and build what the work actually needs.
Keep moving forward
Articles are a starting point. Find the practical tools and next steps that fit where you are in your career.