[{"data":1,"prerenderedAt":3745},["ShallowReactive",2],{"blog-posts":3},[4,253,387,578,815,921,1137,1295,1456,1752,2311,2531,2679,2871,3178,3439,3670],{"id":5,"title":6,"author":7,"body":8,"date":237,"description":238,"extension":239,"image":240,"meta":241,"navigation":242,"path":243,"seo":244,"stem":245,"tags":246,"__hash__":252},"blog\u002Fblog\u002Fagentic-learning-and-the-courses-it-wont-replace.md","Agentic learning, and the courses it will not replace","Peter",{"type":9,"value":10,"toc":225},"minimark",[11,20,23,28,31,34,37,41,44,47,50,54,61,67,73,79,85,91,94,98,105,116,123,126,136,140,143,146,150,157,160,163,181,184,188,191,194,197,201,204,207,210,213],[12,13,14,15,19],"p",{},"A learner types their goal into a chat window. An AI agent takes it from there — diagnosing what they already know, generating an explanation, posing a question, watching the answer, choosing what comes next. No predefined curriculum. No built lessons. The course as an artifact does not exist; the agent ",[16,17,18],"em",{},"is"," the course.",[12,21,22],{},"This kind of agentic learning is real now, for some subjects, with some learners. Whether it should replace structured courses, for everyone and everything, is a more interesting question than the marketing usually allows. The honest answer is that the early evidence is mixed, that the structural problems are not solved, and that the most useful frame is not \"agentic versus structured\" but \"which parts of which learning experience should be each.\"",[24,25,27],"h2",{"id":26},"what-agentic-learning-actually-means","What \"agentic learning\" actually means",[12,29,30],{},"The term gets used loosely, so worth pinning down. By agentic learning I mean a system where an AI agent owns the whole loop: assessing the learner, deciding what to teach next, generating the content, checking whether it stuck, adapting from there. Nothing pre-authored. Nothing version-controlled. The interaction is the entire experience.",[12,32,33],{},"That is meaningfully different from two adjacent things LearnBuilder already does. AI-assisted authoring means an AI helps a human build a course you then deliver — the artifact remains. An AI tutor inside a structured course means a chat helper lives next to fixed lessons — the artifact frames the interaction. Both are useful and both exist today. Pure agentic learning, in the strict sense, is the case where there is no artifact at all.",[12,35,36],{},"This piece is about that strict case.",[24,38,40],{"id":39},"what-an-agent-can-genuinely-do-right-now","What an agent can genuinely do right now",[12,42,43],{},"Conversational explanation that adapts to questions, with patience that no human tutor sustains for free. Practice problem generation in mathematics, language learning, basic programming, and some scientific reasoning. Reasonable inference of a learner's level from a few exchanges. 24\u002F7 availability at near-zero marginal cost per learner. Tutoring quality that is genuinely useful for high-resource domains and major languages — the things the underlying model has seen enough of.",[12,45,46],{},"The most visible production examples are Khan Academy's Khanmigo, which wraps GPT-4 in a constrained tutoring interface, and Duolingo Max, which uses LLMs for conversational language practice and explanation. The largest user population, almost certainly, is informal — students using ChatGPT or Claude as a tutor with no product wrapper around it at all.",[12,48,49],{},"For a curious adult exploring a topic, a child supplementing schoolwork, a developer picking up a library on a Saturday afternoon, this works. The personalisation, the patience, and the price point are all real improvements over the realistic alternative for most learners, which is no tutor at all.",[24,51,53],{"id":52},"what-it-still-cannot-do-and-why-each-gap-matters","What it still cannot do — and why each gap matters",[12,55,56,60],{},[57,58,59],"strong",{},"Reliable assessment of mastery."," The agent can tell whether a learner produced a correct-looking answer in chat. It cannot reliably tell whether the underlying knowledge will transfer to a different context two weeks later. Engagement and learning are not the same thing, and current systems optimise the former by default.",[12,62,63,66],{},[57,64,65],{},"Coherent multi-week curriculum design."," A model can sequence three explanations in a row. Whether the seventh week of a course should revisit week two's concepts, in what form, with what spacing, with which retrieval prompts, is the kind of structural judgement that benefits from intentional design rather than turn-by-turn improvisation.",[12,68,69,72],{},[57,70,71],{},"Verified factual accuracy in specialised or recent domains."," Hallucination in general knowledge has been reduced substantially. In compliance, medicine, law, internal company knowledge, and recently updated regulations, the failure modes are still material and still confident. Retrieval helps; it does not eliminate the problem.",[12,74,75,78],{},[57,76,77],{},"Production-quality interactive media on demand."," Branching scenarios, accurate diagrams, simulations, well-edited video, drag-and-drop activities — these can be generated, but the gap between \"generated\" and \"well-designed\" is large in 2026. For most domains, the well-designed version still takes a human deciding.",[12,80,81,84],{},[57,82,83],{},"An auditable trail of what was taught."," Compliance is the obvious case, but professional certification, regulated industries, accreditation, and most workplace L&D need a thing to point to that says: this is the content, this is the version, these are the outcomes, here is the evidence learners met them. A pure agentic system, by design, does not produce that artifact.",[12,86,87,90],{},[57,88,89],{},"Cohort dynamics."," Group discussion, shared deadlines, peer-explanation effects, the social motivation of a course running in real time with other humans on the same schedule. An agent can produce content. It does not produce a class.",[12,92,93],{},"Each of these gaps is structural, not a temporary engineering limitation that closes in eighteen months. Some will improve materially. Others — particularly the artifact and the cohort — are not really problems the agentic frame is trying to solve.",[24,95,97],{"id":96},"what-the-research-actually-says","What the research actually says",[12,99,100,101,104],{},"The headline number people quote in favour of AI tutoring is Bloom's \"2 sigma\" finding from 1984 — that one-on-one human tutoring produced learning gains roughly two standard deviations above a conventional classroom. The implied argument is that AI agents can replicate that economically. The trouble is that the original study is about ",[16,102,103],{},"human"," tutors, and the empirical record for LLM tutors is genuinely mixed.",[12,106,107,108,111,112,115],{},"The single most useful study to know about is Bastani et al. (2024) at Wharton, ",[16,109,110],{},"Generative AI Can Harm Learning",". They ran a controlled experiment in a high-school mathematics context. Students with unrestricted access to GPT-4 during practice scored ",[16,113,114],{},"worse"," on a follow-up exam than students without GPT-4 access at all. A constrained \"tutor mode\" version — which guided students rather than solving for them — performed better. The mechanism is the part that matters: a too-helpful agent removes the productive struggle that builds durable knowledge. Practice without resistance does not generate retention.",[12,117,118,119,122],{},"This connects to long-established cognitive science. Robert Bjork's work on \"desirable difficulties\" predicts that frictionless explanations, given before a learner has wrestled with a problem, reduce retention. John Sweller's cognitive load theory predicts something compatible — that ",[16,120,121],{},"too much"," scaffolding can be as harmful as too little. Agentic systems optimised for user satisfaction will lean toward frictionless, because that is what learners ask for in the moment. That is exactly the failure mode the research warns about.",[12,124,125],{},"On the other side of the ledger, there are early reports — not yet a robust RCT base — that constrained AI tutors with explicit pedagogical guardrails (Khanmigo is the example most often cited) modestly help with engagement and self-paced practice. Ethan Mollick and colleagues at Wharton have written about AI co-tutoring in classroom contexts producing reasonable results when the agent is configured to coach rather than solve.",[12,127,128,129,132,133,135],{},"The honest summary: the technology is improving fast, the early research is small-N and short-horizon, and the field has not yet shown durable, transferable learning gains from pure agentic systems comparable to what Bloom's 2-sigma claim promises. If you read one paper before adopting agentic learning at scale, make it the Bastani study. The finding that an ",[16,130,131],{},"unguided"," AI tutor can produce ",[16,134,114],{}," outcomes than no tutor at all is the thing most marketing in this space quietly ignores.",[24,137,139],{"id":138},"what-this-article-is-probably-getting-wrong","What this article is probably getting wrong",[12,141,142],{},"I am cautious because the current evidence is mixed and the structural problems — assessment, traceability, productive struggle — are not solved. But agentic systems are improving quickly. Bastani used GPT-4 in 2023. A 2026 system with better retrieval, stronger guardrails, real assessment hooks, and explicit pedagogical scaffolding could produce a very different result. A future where pure agentic learning genuinely matches structured courses for a much broader range of contexts is plausible, and it could arrive faster than this piece implies.",[12,144,145],{},"Treat the \"not ready to replace\" claim as load-bearing on today's evidence. If the evidence shifts, the claim shifts. The structural points about the artifact and the cohort survive technological progress; the empirical points about learning outcomes do not necessarily.",[24,147,149],{"id":148},"why-structured-courses-survive-where-the-artifact-matters","Why structured courses survive where the artifact matters",[12,151,152,153,156],{},"A course is a ",[16,154,155],{},"durable artifact",". It can be versioned, reviewed, signed off by a compliance officer or subject expert, branded, translated, deployed across an LMS, and completed by a learner with a record that completion happened. These are not nice-to-have features. For workplace training, professional certification, regulated industries, and any context where what was taught matters to someone other than the learner, the artifact is the point.",[12,158,159],{},"Agentic learning, by design, does not produce that artifact. The conversation happened. The learner improved, or did not. There is no canonical statement of \"this is what the course taught\" that an auditor or accreditor can examine.",[12,161,162],{},"So the structural argument is straightforward: built courses survive wherever the artifact matters. That is most workplace L&D, most compliance, most certification, most regulated content, most cohort programmes, and most situations where a stakeholder downstream of the learner has a reason to care what was actually covered.",[12,164,165,166,169,170,175,176,180],{},"What does ",[16,167,168],{},"not"," survive untouched is the assumption that a structured course must be inert. LearnBuilder is already moving toward courses that have agentic capabilities inside them — AI course generation as a drafting tool (",[171,172,174],"a",{"href":173},"\u002Fblog\u002Fai-assisted-hand-built-or-both-learnbuilder","covered in detail in this earlier post","), AI tutoring inside lessons, AI-generated practice variants, AI-translated subtitles across sixty languages (",[171,177,179],{"href":178},"\u002Fblog\u002Flearnbuilder-april-2026-release","from the April release","), AI-generated interactive blocks. The structural unit is the course; the texture inside it is increasingly agentic.",[12,182,183],{},"The bet is hybrid: built courses with rich agentic features inside, not pure agentic loops replacing the course as a concept. Replacement everywhere is unlikely. Replacement in the artifact-mattering contexts is structurally blocked, not just empirically weak.",[24,185,187],{"id":186},"where-pure-agentic-learning-probably-wins","Where pure agentic learning probably wins",[12,189,190],{},"The contexts where the artifact does not matter are real, and large, and growing. A curious adult learning a topic for themselves. A child supplementing schoolwork. A developer picking up a new library on a weekend. A student preparing for a standardised exam. Conversational language practice. Hobbyist skill building. Self-directed exploration of any kind where the learner is the only meaningful stakeholder.",[12,192,193],{},"In those contexts the agentic loop is genuinely powerful, and a built course looks slow and clumsy by comparison. The future of self-directed learning probably is largely agentic, and that is good news — the economics of patient, infinitely-available tutoring at near-zero marginal cost change what is possible for learners who would otherwise have nothing.",[12,195,196],{},"The pure-agentic future is real. It is just not the whole future.",[24,198,200],{"id":199},"the-useful-question-is-which-parts-not-which-side","The useful question is which parts, not which side",[12,202,203],{},"The framing of \"agentic replaces structured\" is a marketing artefact, not a useful design question. The useful question is which parts of a learner's experience should be agentic and which should be structured, and the answer changes with the stakes, the audience, the accountability, and the evidence the learning needs to have produced.",[12,205,206],{},"For self-directed exploration: lean agentic. For certification and compliance: lean structured. For most workplace and educational learning, where neither extreme fits cleanly: build a structured course and put agentic capabilities inside it. The course gives you the artifact, the audit trail, the cohort, the brand, and the version control. The agentic features give you the personalisation, the patience, the on-demand practice, and the explanation tailored to the learner in front of you.",[12,208,209],{},"The research, on current evidence, suggests the gains live in that hybrid space rather than at either pole. The structural arguments suggest the same. Build for both, and revisit the balance as the evidence moves.",[211,212],"hr",{},[12,214,215],{},[16,216,217,218,224],{},"If you want to see hybrid in practice — built courses with AI tutoring, AI-generated practice, and AI-translated subtitles inside — the ",[171,219,223],{"href":220,"rel":221},"https:\u002F\u002Flearnbuilder.org",[222],"nofollow","free trial"," is the place to start.",{"title":226,"searchDepth":227,"depth":227,"links":228},"",2,[229,230,231,232,233,234,235,236],{"id":26,"depth":227,"text":27},{"id":39,"depth":227,"text":40},{"id":52,"depth":227,"text":53},{"id":96,"depth":227,"text":97},{"id":138,"depth":227,"text":139},{"id":148,"depth":227,"text":149},{"id":186,"depth":227,"text":187},{"id":199,"depth":227,"text":200},"2026-05-24","An honest look at agentic learning — AI agents that design and deliver the whole learning experience on the fly. What is genuinely possible today, what early research like Bastani et al.'s 'Generative AI Can Harm Learning' actually shows, and which parts of structured course design no agent has yet replaced.","md","\u002Fblog\u002Fagentic-learning-and-the-courses-it-wont-replace.webp",{},true,"\u002Fblog\u002Fagentic-learning-and-the-courses-it-wont-replace",{"title":6,"description":238},"blog\u002Fagentic-learning-and-the-courses-it-wont-replace",[247,248,249,250,251],"AI","agentic learning","research","instructional design","course authoring","5XuMXcLRdIYmz1JuQCPWEIdAcZVwbyqxXeWAwxed3TA",{"id":254,"title":255,"author":7,"body":256,"date":377,"description":378,"extension":239,"image":379,"meta":380,"navigation":242,"path":381,"seo":382,"stem":383,"tags":384,"__hash__":386},"blog\u002Fblog\u002Fbuilding-a-course-by-hand-on-purpose.md","Building a course by hand, on purpose",{"type":9,"value":257,"toc":369},[258,261,264,268,271,274,277,281,284,287,290,294,297,304,315,318,322,325,332,335,339,342,345,348,352,355,358,360],[12,259,260],{},"For the last few months, almost every course I made started the same way: upload a document, pick an instructional approach, wait, refine. It works, and I am not going to pretend otherwise. Then this week I built a course the slow way — from an empty lesson, one block at a time — and I noticed how much I had stopped doing.",[12,262,263],{},"Not stopped doing badly. Stopped doing at all.",[24,265,267],{"id":266},"what-the-generated-workflow-gets-right","What the generated workflow gets right",[12,269,270],{},"The blank lesson is the hardest part of a course. Not the writing, not the questions — the structure. Where does this start, what are the lessons even called, what belongs in lesson three versus lesson seven. An experienced designer can do this, but it still takes a focused hour or two of staring before the shape appears.",[12,272,273],{},"AI course generation hands you that shape. Upload a forty-page policy document, choose whether you want an action-first or objective-first structure, and you get a complete scaffold with explanatory content, knowledge checks, and activities already sequenced. When the deadline is Friday and the source is a dense, well-organised document, this is not a shortcut. It is the only realistic way to ship on time, and I have shipped courses I am happy with this way.",[12,275,276],{},"So this is not a piece about AI being bad at course-making. It is a piece about what the fast path skips over, and why that is sometimes worth slowing down for.",[24,278,280],{"id":279},"the-loop-that-generation-skips","The loop that generation skips",[12,282,283],{},"When you build a course by hand, you spend most of your time inside a loop. You add an interaction. You look at it. It is not broken — it just is not right. You delete it. You try a different way into the same idea. You sit with a question for a while, wondering whether it is actually testing anything or just checking that someone scrolled past the previous screen. You move a block, dislike it, move it back.",[12,285,286],{},"That loop has a lot of dead time in it, and the dead time is where the design decisions actually happen. Trying something and rejecting it is not wasted motion — it is how you find out what you think. By the time a hand-built lesson is finished, you have seen and discarded five versions of it, and the one that survived is better for having competition.",[12,288,289],{},"A generated course arrives without that loop ever having run. The decisions were still made — they were just made statistically, by the model, in one pass, and you never got to see the alternatives it discarded. Often the result is fine. But \"fine\" is exactly the word, and you cannot always tell whether a generated lesson is the best version of itself or simply the most probable one.",[24,291,293],{"id":292},"humour-and-specifics-dont-survive-averaging","Humour and specifics don't survive averaging",[12,295,296],{},"The thing I missed most is small: the jokes.",[12,298,299,300,303],{},"A joke in a course works because it is specific. It lands because it refers to this audience, this office, this piece of software everyone secretly hates, this running gag the team will recognise. AI-generated humour tends to be the ",[16,301,302],{},"average"," joke — safe, mildly amusing, equally applicable to any audience anywhere, which is another way of saying it connects with no one in particular.",[12,305,306,307,310,311,314],{},"The same is true of examples. The example that makes an abstract point finally click is usually the oddly specific one — the slightly embarrassing scenario, the thing that went wrong last Tuesday, the detail too particular for a model to have reached for. A model reaches for the ",[16,308,309],{},"representative"," example, because that is what it is built to do. A person reaches for the ",[16,312,313],{},"memorable"," one. Those are not the same example, and a learner can feel the difference even if they could not name it.",[12,316,317],{},"This is an observation, not a research finding. But after building a course by hand, I think the human touch in learning content is mostly this: a steady stream of small, specific choices that no average would ever produce.",[24,319,321],{"id":320},"editing-a-generated-course-is-real-work","Editing a generated course is real work",[12,323,324],{},"The other thing I had half-forgotten is how much work it is to amend a generated course into the one you actually wanted.",[12,326,327,328,331],{},"The demo always shows generation as the whole task — document in, course out. In practice the generated draft is the ",[16,329,330],{},"start"," of the task. Sometimes reshaping it takes nearly as long as building from scratch: rewriting questions that test recall when they should test judgement, resequencing lessons, cutting a section the model was confident belonged. And it is a particular kind of slow, because you are also reverse-engineering decisions you did not make and quietly arguing with them.",[12,333,334],{},"Building by hand, there is nothing to undo. Every decision is yours from the first block, so the work only ever moves forward. Neither path is free. They are just different shapes of work, and it is worth knowing which shape you are signing up for before you start.",[24,336,338],{"id":337},"im-not-romanticising-the-slow-way","I'm not romanticising the slow way",[12,340,341],{},"The slow loop is also where procrastination hides. Not every interaction I deleted and retried this week was design judgement — some of it was me fiddling because fiddling is pleasant, and a deadline was comfortably far away. The craft is enjoyable, and enjoyable work is easy to mistake for important work.",[12,343,344],{},"Manual authoring is genuinely slower, and for a great deal of content the craft does not matter. A compliance refresher covering a regulation that changed one clause does not need my jokes or my oddly specific examples. It needs to be accurate, clear, and finished. If I hand-built every course, most of them would ship late and most would be no better for the extra time. Pleasure is not the same as value, and confusing the two is how a project quietly slips a month.",[12,346,347],{},"So the honest position is not \"build by hand.\" It is \"know which courses deserve it.\"",[24,349,351],{"id":350},"the-point-is-the-dial-not-the-verdict","The point is the dial, not the verdict",[12,353,354],{},"LearnBuilder was built so you do not have to settle this argument once and for all. You can generate a full course from a document. You can open an empty lesson and build every block yourself. Or — most usefully — you can generate the scaffold and then spend your craft only where it earns its keep: rewrite the three questions the course actually turns on, hand-make the one branching scenario that matters, add the joke that this specific audience will get, and leave the rest as generated.",[12,356,357],{},"The dial moves per course. Honestly, it moves per mood and per deadline too. This week I wanted the slow loop, so I took it, and the course is better for it. Next week, with a policy document and a Friday, I will generate the whole thing and feel no guilt at all. A tool that only does one of those is making that choice on your behalf — and it is a choice that should stay with the person who knows the audience, the subject, and how much time they actually have.",[211,359],{},[12,361,362],{},[16,363,364,365,368],{},"If you want to feel the difference yourself, the ",[171,366,223],{"href":220,"rel":367},[222]," lets you generate a course and build one by hand in the same afternoon.",{"title":226,"searchDepth":227,"depth":227,"links":370},[371,372,373,374,375,376],{"id":266,"depth":227,"text":267},{"id":279,"depth":227,"text":280},{"id":292,"depth":227,"text":293},{"id":320,"depth":227,"text":321},{"id":337,"depth":227,"text":338},{"id":350,"depth":227,"text":351},"2026-05-15","After months of generating courses with AI, I built one by hand again — and noticed everything the fast workflow quietly removes: the loop of trying and deleting, the specific jokes, the human touch. Why LearnBuilder lets you set how much AI you want, per course.","\u002Fblog\u002Fbuilding-a-course-by-hand-on-purpose.webp",{},"\u002Fblog\u002Fbuilding-a-course-by-hand-on-purpose",{"title":255,"description":378},"blog\u002Fbuilding-a-course-by-hand-on-purpose",[251,247,250,385],"manual authoring","_k7gtaAxKVXj9i_0rjsWewSXXrJXb-2Xewwpfy9fPA4",{"id":388,"title":389,"author":7,"body":390,"date":565,"description":566,"extension":239,"image":567,"meta":568,"navigation":242,"path":178,"seo":569,"stem":570,"tags":571,"__hash__":577},"blog\u002Fblog\u002Flearnbuilder-april-2026-release.md","What shipped in LearnBuilder this week: a calmer editor, smarter AI course creation, and 60 languages",{"type":9,"value":391,"toc":554},[392,395,398,402,405,408,411,415,418,421,424,428,431,434,438,441,444,447,451,454,482,485,489,500,503,506,510,513,516,520,523,526,530,533,536,544,546],[12,393,394],{},"This update is bigger than the usual weekly patch. A round of editor cleanup, several new layout primitives, two upgrades to the interactive slideshow builder, a meaningful expansion of how AI course creation handles instructional design and audience, full localisation across sixty languages, AI-translated video subtitles, and finer control over AI video generation. Most of it has been on the roadmap for a while; this is the week it landed together.",[12,396,397],{},"Below is what changed and, more importantly, what each change actually unlocks for the people using the tool day to day.",[24,399,401],{"id":400},"a-calmer-editor","A calmer editor",[12,403,404],{},"The editor has been cleaned up. Toolbars that were always visible are now visible when you need them. Controls that overlapped with the content area have been moved out of the way. The default state when you open a lesson is closer to the published view, with the chrome receding until you start editing.",[12,406,407],{},"The benefit is less obvious in screenshots than in practice. For people who spend hours a day in the editor, removing visual noise is the difference between an interface that feels heavy and one that gets out of the way. It also makes review sessions easier — when you walk a stakeholder through a draft, they see the lesson, not the editing interface.",[12,409,410],{},"This is the kind of change that does not have a feature name. It is also one of the most-requested changes from the early users, so it is worth calling out explicitly.",[24,412,414],{"id":413},"block-backgrounds-full-width-layouts-and-proper-spacing-controls","Block backgrounds, full-width layouts, and proper spacing controls",[12,416,417],{},"Blocks now support backgrounds — solid colours, images, or gradients — and a full set of margin and padding controls. They can also stretch to 100% width of the lesson, breaking out of the central content column when the design calls for it.",[12,419,420],{},"In aggregate, this is the difference between content that looks like a default theme and content that looks like yours. A full-width image block with overlay text. A section break with a coloured background that signals a change of pace. A scenario set against a gradient that matches the topic. A breathing-room margin around a quiz so it does not feel jammed against the explanatory text above.",[12,422,423],{},"Brand colours and fonts are still configured at the account level and inherited automatically. The new controls give designers control over the per-block visual rhythm of a lesson without breaking the system.",[24,425,427],{"id":426},"full-screen-mode-in-the-interactive-slideshow-builder","Full-screen mode in the interactive slideshow builder",[12,429,430],{},"The interactive slideshow builder now has a full-screen editing mode. The rest of the application chrome collapses, the canvas expands, and you get the screen space back that complex interactions actually need.",[12,432,433],{},"This matters more than it sounds. Layered interactions — a branching scenario with multiple pathways, a simulation with three or four interactive elements, a custom layout with a dozen tracked components — were genuinely cramped in the standard editor. Full-screen mode gives you the room to lay them out without zooming in and out, and it keeps element selection and the property panel where you expect them. For anyone building anything beyond a one-element interaction, this is a quality-of-life change with real consequences for how much you can fit on a slide and still keep editing.",[24,435,437],{"id":436},"an-svg-element-with-editable-source-code","An SVG element with editable source code",[12,439,440],{},"Interactive slides now include an SVG element. Drop one in, paste the SVG markup, and the source is editable directly inside the element panel.",[12,442,443],{},"The reason this is in the product: AI tools and designer toolchains both produce SVG natively. An icon library, a custom diagram, an org chart, a process flow, a visual you generated with an external tool — all of these arrive as SVG. Until now, getting them into a lesson meant exporting to PNG, losing scalability, and editing the source elsewhere. Now you can paste it in and tweak it in place, change a colour, add an animation, fix a label. For technical designers, this is also the cleanest way to drop in a hand-built SVG that is reused across lessons.",[12,445,446],{},"This pairs with the existing AI coding block. SVG covers the static-but-editable visual case; the AI coding block covers the dynamic-interaction case.",[24,448,450],{"id":449},"four-instructional-design-approaches-for-ai-course-creation","Four instructional-design approaches for AI course creation",[12,452,453],{},"AI course creation now lets you choose between four instructional-design approaches when generating a course outline:",[455,456,457,464,470,476],"ul",{},[458,459,460,463],"li",{},[57,461,462],{},"Standard"," — a balanced structure with explanatory content, knowledge checks, and scenario activities sequenced toward the learning objectives.",[458,465,466,469],{},[57,467,468],{},"Action-first"," — opens each lesson with the task or behaviour the learner needs to perform, then fills in the knowledge required to do it. Good for skills training and procedural content.",[458,471,472,475],{},[57,473,474],{},"Objective-first"," — leads with the learning objective and structures content around evidence that the objective has been met. Good for compliance, certification, and assessment-driven programmes.",[458,477,478,481],{},[57,479,480],{},"Story-driven"," — frames the content around a narrative or scenario that unfolds across the lesson. Good for soft-skills training, ethics, and topics where context matters more than information transfer.",[12,483,484],{},"The benefit is that the AI is no longer making one set of structural assumptions on your behalf. The same source material — a policy document, a product spec, a subject-matter brief — produces meaningfully different courses depending on the approach. Designers who know which approach fits the audience can pick it directly, instead of regenerating with prompt tweaks until the structure feels right.",[24,486,488],{"id":487},"experience-level-and-learning-context","Experience level and learning context",[12,490,491,492,495,496,499],{},"In the same step where you choose the instructional approach, you can now set the ",[57,493,494],{},"experience level"," of the target audience (novice, intermediate, advanced) and the ",[57,497,498],{},"learning context"," in which the course will be used (onboarding, compliance, performance support, deep-skill development, refresher, and others).",[12,501,502],{},"This shifts what the AI generates in concrete ways. A novice-onboarding course gets more scaffolding, more retrieval practice, and shorter cognitive-load chunks. An advanced performance-support course skips foundational explanations and goes straight to decision support and edge cases. A compliance refresher generates differently from an initial compliance course, even on the same regulation.",[12,504,505],{},"The combination of approach + experience + context is what an instructional designer is doing in their head when they brief a developer. Surfacing it as explicit input means the first draft is closer to what the designer would have asked for, with fewer rounds of regeneration.",[24,507,509],{"id":508},"sixty-languages-with-course-translation","Sixty languages, with course translation",[12,511,512],{},"Course language selection now covers sixty languages, and full course translation between any of them is available with one action. Generated content respects the conventions of the target language — not just word-for-word translation, but appropriate idioms, examples, and instructional patterns.",[12,514,515],{},"For organisations with regional teams, this collapses what used to be a multi-week localisation pipeline into something a single designer can run from the editor. The translated course retains its structure, its branding, its interactive elements, and its accessibility metadata. You still want a human review pass for any high-stakes content, especially compliance, but the starting point is far closer to publish-ready than a machine translation of HTML output would be.",[24,517,519],{"id":518},"ai-translated-video-subtitles","AI-translated video subtitles",[12,521,522],{},"Video subtitles can now be AI-translated alongside the course translation. Upload a video once with subtitles in the source language, and translated versions are produced for every other language the course is published in.",[12,524,525],{},"This is the second piece of the localisation story. Translated text without translated subtitles produces a course that is bilingual in the worst way — the lesson is in Spanish, the video is captioned in English. The subtitle translation closes that gap. As with course translation, a human pass is recommended for video where the script carries weight.",[24,527,529],{"id":528},"start-and-end-frame-control-for-ai-video-generation","Start and end frame control for AI video generation",[12,531,532],{},"AI video generation now accepts a start frame and an end frame, in addition to the prompt. You upload an image of where the video should begin, an image of where it should end, and the model generates the motion between them.",[12,534,535],{},"The benefit is control over output that previously required prompt-and-pray iteration. For product walkthroughs, scenario reconstructions, or any video where the visual continuity with the rest of the lesson matters, you can now anchor the result to specific frames you have control over. The animation can pick up from a still you placed earlier in the lesson and end on a still you use afterwards. It is not the same as a hand-edited cut, but for the use cases AI video is realistically good for, this closes most of the gap between \"approximately what I wanted\" and \"what I wanted.\"",[12,537,538,539,543],{},"If you come across anything in this release that does not behave the way you expect or have ideas for new features, ",[171,540,542],{"href":541},"mailto:peter@learnbuilder.org","contact us",".",[211,545],{},[12,547,548],{},[16,549,550,551,224],{},"The new features are live for all accounts. If you want to try them on a real course, the ",[171,552,223],{"href":220,"rel":553},[222],{"title":226,"searchDepth":227,"depth":227,"links":555},[556,557,558,559,560,561,562,563,564],{"id":400,"depth":227,"text":401},{"id":413,"depth":227,"text":414},{"id":426,"depth":227,"text":427},{"id":436,"depth":227,"text":437},{"id":449,"depth":227,"text":450},{"id":487,"depth":227,"text":488},{"id":508,"depth":227,"text":509},{"id":518,"depth":227,"text":519},{"id":528,"depth":227,"text":529},"2026-04-30","Block backgrounds and full-width layouts, full-screen interactive slides, an editable SVG element, four instructional-design approaches for AI course creation, audience-aware tailoring, 60 course languages with translation, AI-translated video subtitles, and start\u002Fend frame control for video generation.","\u002Fblog\u002Flearnbuilder-april-2026-release.webp",{},{"title":389,"description":566},"blog\u002Flearnbuilder-april-2026-release",[572,573,247,574,575,576],"LearnBuilder","release notes","interactive slides","video","localisation","u8RcocqVF8BKz62WvCy6dNw1lqtqoTGm10sLxCXx6Wg",{"id":579,"title":580,"author":7,"body":581,"date":805,"description":806,"extension":239,"image":807,"meta":808,"navigation":242,"path":809,"seo":810,"stem":811,"tags":812,"__hash__":814},"blog\u002Fblog\u002Fwill-claude-design-replace-elearning-authoring.md","Will Claude Design replace e-learning authoring tools?",{"type":9,"value":582,"toc":796},[583,591,594,601,605,608,611,617,621,624,630,636,642,645,649,652,658,669,675,681,687,693,696,700,703,706,709,712,715,719,722,733,740,743,746,749,753,756,759,762,766,769,772,775,777],[12,584,585,590],{},[171,586,589],{"href":587,"rel":588},"https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fclaude-design-anthropic-labs",[222],"Claude Design"," shipped into Anthropic Labs on April 17. It is a real product, not a chat-window improvisation. You describe a design — a prototype, a wireframe, a pitch deck, a marketing asset — and it generates one inside a dedicated visual-collaboration interface running on Claude Opus 4.7. You can leave inline comments, edit elements directly, nudge things with custom adjustment sliders, apply your organization's design system automatically, and export to Canva, PDF, PPTX, or HTML. Designs hand off to Claude Code when they need to become real software.",[12,592,593],{},"The first time you see it generate a working interactive prototype from a paragraph of description, the natural question follows quickly. If Claude Design can do this, why does anyone still pay for Articulate Storyline, Adobe Captivate, iSpring, Rise, or any of the other authoring tools that sit at the centre of the standard L&D toolchain? The first impression makes the answer feel obvious: they don't, not for long.",[12,595,596,597,600],{},"I have been building ",[171,598,572],{"href":220,"rel":599},[222]," for the last year, so I have a stake in this question. I have also spent enough time pushing the limits of what generated artifacts can actually do in a course context to know that the easy answer is wrong in both directions. Claude Design is genuinely going to absorb work that authoring tools used to handle, especially the visual-design and prototyping side. It is also pointed at a different problem than e-learning, and the parts it does not solve are the parts L&D teams cannot ship without.",[24,602,604],{"id":603},"what-claude-design-actually-is-and-is-not","What Claude Design actually is — and is not",[12,606,607],{},"It is worth being precise about the product, because a lot of the discussion around it has blurred into a general \"AI replaces design tools\" frame that misses what Anthropic actually built.",[12,609,610],{},"Claude Design is positioned for designers, product managers, founders, marketers, and non-designers who need to produce visual material. The output categories listed in the launch are realistic prototypes, product wireframes, design explorations, pitch decks, and marketing collateral, with code-powered prototypes that can include voice, video, and 3D. The handoff path to engineering is Claude Code. The collaboration model is organization-scoped sharing with design-system enforcement.",[12,612,613,614,616],{},"What it is ",[16,615,168],{}," is an e-learning tool. There is no concept of a learning objective, a knowledge check, a retrieval-practice activity, a branching scenario tied to a competency, a learner state that survives across sessions, a SCORM or xAPI export, or an LMS integration. None of this is a criticism — Claude Design is not trying to do those things. It is the e-learning gap, and it is the gap that matters when you ask whether the tool replaces an authoring platform.",[24,618,620],{"id":619},"what-claude-design-does-genuinely-well-and-undermines-about-traditional-authoring-tools","What Claude Design does genuinely well — and undermines about traditional authoring tools",[12,622,623],{},"Three pieces of Claude Design hit at the heart of what most authoring tools used to charge for.",[12,625,626,629],{},[57,627,628],{},"Generation quality at the surface."," A prototype that runs in the browser, looks competent, and demonstrates the interaction logic correctly — produced from a paragraph of description in well under a minute. For the kind of one-off visual artifact that a designer used to spend two days on, this is a different category of work, not a marginal speed-up.",[12,631,632,635],{},[57,633,634],{},"Direct editing of generated output, not re-prompting."," This is the piece that surprised me. The first wave of generated-artifact tools forced you to nudge by re-describing what you wanted, which usually triggered a partial regeneration, which usually changed things you did not want changed. Claude Design exposes the elements directly. Inline comments, direct manipulation, adjustment sliders — you fix the thing yourself instead of gambling on the next round. That closes the iteration problem that breaks the chat-window workflow.",[12,637,638,641],{},[57,639,640],{},"Design systems applied automatically."," Brand consistency was a real moat for authoring tools — your Storyline output looked like your Storyline output because the template enforced it. Claude Design folds your organization's design system into generation directly, so the artifact lands inside your visual identity from the first draft. Combined with org-scoped sharing, the consistency-and-collaboration argument that traditional template tools used to make is now table stakes in a different product.",[12,643,644],{},"If you are an L&D team building a single high-craft prototype, a marketing asset for a learning campaign, a pitch deck for a programme, or an exploratory wireframe — Claude Design is now in the toolkit, and it should be. The question is whether that is the same as the work an authoring tool actually does.",[24,646,648],{"id":647},"what-e-learning-needs-that-visual-design-does-not","What e-learning needs that visual design does not",[12,650,651],{},"Here is where the analysis splits. Claude Design is a visual collaboration tool. E-learning has visual-collaboration needs, but the visible surface of an interaction is maybe a third of the work. The rest is the part that does not show up in a prototype review.",[12,653,654,657],{},[57,655,656],{},"Instructional design as a constraint, not a style."," Learning content is structured by learning objectives, sequenced for retrieval and transfer, balanced across content types so a learner is not reading text blocks for twenty minutes. A general visual tool has no model of any of this. It will produce a beautiful pitch deck on photosynthesis. It will not insist that the deck include knowledge checks, that the knowledge checks test recall before recognition, that the scenario activity is anchored to a measurable behaviour, or that the lesson does not exceed cognitive-load thresholds. That is the instructional-design layer, and it is the layer that decides whether the content actually teaches.",[12,659,660,663,664,668],{},[57,661,662],{},"Persistence of learner state."," A learner who closes a tab in the middle of a branching scenario should resume where they left off. A learner who spent fifteen minutes building a configuration in a simulation should find it on return. Generated prototypes do not get session persistence as a default, and there is no equivalent of ",[665,666,667],"code",{},"cmi.suspend_data"," in a Claude Design export.",[12,670,671,674],{},[57,672,673],{},"Reporting and grading."," For graded interactions, the platform needs to know what the learner did. Did they reach the optimal configuration? Did they explore the full range, or click once and leave? Did they choose the best answer, the second-best, or the trap option? A hand-built interaction wires this up deliberately. A generated prototype does not have a contract for what to report — and there is no LMS at the other end to receive it anyway.",[12,676,677,680],{},[57,678,679],{},"SCORM, xAPI, cmi5, AICC."," Corporate L&D runs on tracking, and tracking only works when the content reports it correctly through the integration format the LMS speaks. Export to PDF, PPTX, HTML, or Canva — all useful — does not produce a SCORM package or an xAPI statement. This is unglamorous plumbing, and it is the difference between content you can demo and content you can ship into a corporate compliance programme.",[12,682,683,686],{},[57,684,685],{},"Accessibility maintained over a course, not at the prototype level."," WCAG AA across forty lessons is a sustained property, not a one-time review. Missing alt text, hover-only interactions with no keyboard equivalent, missing captions, contrast failures — these are easy to introduce and easy to miss when the focus is on the interaction. A learning platform needs ongoing accessibility checking; a design tool checks at the point of design.",[12,688,689,692],{},[57,690,691],{},"Course-scale coherence."," A pitch deck is one artifact. A course is fifteen lessons that share an instructional spine, sequence toward a competency, share a visual identity, share a glossary of terms, share a tone, and report into a single learner record. Even with a strong design system, generating each lesson independently in a visual tool produces fifteen pretty artifacts with no spine connecting them.",[12,694,695],{},"None of these gaps are weaknesses of Claude Design. They are simply outside the problem it set out to solve.",[24,697,699],{"id":698},"where-this-leaves-authoring-tools","Where this leaves authoring tools",[12,701,702],{},"Two things are happening at once, and they look contradictory until you separate them.",[12,704,705],{},"The first: traditional template-driven authoring tools are in real trouble. Tools whose value proposition was \"we built a drag-and-drop interaction library so you don't have to code, and our templates enforce consistency\" are now competing with a model that generates the interaction, applies your design system, and lets you edit directly. The interaction library was the moat. Generation dissolves it. The tools whose pricing pages still emphasise \"200+ ready-made templates\" are about to find that this is not the value any more.",[12,707,708],{},"The second: the e-learning-specific layer — instructional design opinions, learning-objective scaffolding, knowledge checks generated alongside content, SCORM and xAPI export, LMS integration, learner-state persistence, course-scale coherence, ongoing accessibility — is not solved by a general visual design tool, because it was never the same problem.",[12,710,711],{},"The authoring tools that survive will be the ones that absorb generation rather than resist it, and that focus on the L&D-specific layer that a general tool will not build. That means a different shape of product: AI inside the editor rather than a separate workflow you copy-paste from. Generated interactions that are editable, not opaque. The plumbing — tracking, persistence, accessibility, reporting, branding, collaboration, LMS export — handled by the platform so the designer can focus on the learning design.",[12,713,714],{},"This is what LearnBuilder is built around. The piece I want to talk about specifically is the interactive slideshow builder, because it is where generation and editing meet most directly — and it is the closest thing in an e-learning tool to what Claude Design does for visual design.",[24,716,718],{"id":717},"editing-ai-created-interactions-in-the-learnbuilder-slideshow-builder","Editing AI-created interactions in the LearnBuilder slideshow builder",[12,720,721],{},"The LearnBuilder slideshow builder shares a core idea with Claude Design: generated output should be editable in the same surface as hand-built output, not trapped behind a re-prompt loop.",[12,723,724,725,728,729,732],{},"When the AI generates an interaction inside a LearnBuilder slide, it does not produce one opaque blob. It creates elements — shapes, animations, bindings, text, images, controls — in the same element tree you build with by hand. After generation, you can select the rotor on a windmill simulation and change its colour. You can grab the slider and reposition it. You can resize the base, change the typography of a label, tweak the easing of an animation. The same controls that work for hand-built elements work for AI-built ones, because they are the same elements. This is the difference between ",[16,726,727],{},"editable"," and ",[16,730,731],{},"re-promptable",": you fix the thing yourself, with direct manipulation, in seconds, without spending tokens or waiting for a regeneration.",[12,734,735,736,739],{},"The second piece is element-scoped AI. You can select a single element — ",[16,737,738],{},"this rotor"," — and ask the AI to modify just that. \"Spin more slowly.\" \"Change the pivot to the top of the pole.\" The model gets scoped context, not the whole slide. Tokens stay low. The rest of the slide is guaranteed not to change because the AI literally cannot touch it.",[12,741,742],{},"The third piece is exposed code. When the AI writes JavaScript for a custom interaction, that code is saved as a custom JS interaction on the slide. It is not hidden behind an opaque \"AI layer.\" If the AI gets you eighty percent of the way, a developer on your team can open the code, read it, and cover the last twenty. The same hatch is available without AI: hand-write the JS for an interaction that has no template, paste it in, and it runs inside a properly structured course with all the surrounding infrastructure handled by the platform.",[12,744,745],{},"This is the same product shape Claude Design uses for visual design: generation, direct editing, element-level scoping, and a hatch to code. The difference is everything around it. In LearnBuilder, the generated interaction lives inside a lesson that lives inside a course that exports as SCORM 1.2, SCORM 2004, xAPI, or cmi5; that respects an account-level brand configuration applied to every lesson; that is checked by a built-in accessibility scanner before publish; that supports real-time multi-author collaboration; that tracks learner progress and reports completion to the LMS; that persists learner state across sessions; and that generates retrieval questions and scenario activities alongside explanatory content because the AI is working from instructional principles, not visual prompts.",[12,747,748],{},"Generation is the same idea. The platform around it is what makes it e-learning rather than a prototype.",[24,750,752],{"id":751},"where-claude-design-genuinely-wins-for-ld","Where Claude Design genuinely wins for L&D",[12,754,755],{},"The honest caveat. Even within an L&D team, there are jobs Claude Design is now plainly the right tool for.",[12,757,758],{},"Pitch decks for new programmes. Marketing assets for a course launch. Internal wireframes for a new learner experience that has not yet been built. Stakeholder prototypes that need to be polished and on-brand without engineering involvement. High-craft one-off interactions where a designer wants total control over the visual language and is prepared to handle the integration manually. The handoff path to Claude Code is also genuinely useful when an interaction needs to become production code rather than stay a prototype.",[12,760,761],{},"For the work of producing courses — the day-job of an instructional designer in a corporate L&D function — Claude Design is not the tool, because it is not trying to be. The right combination is Claude Design for the visual-design adjacent work and a generation-native authoring tool for the courses themselves.",[24,763,765],{"id":764},"replacement-is-the-wrong-frame","Replacement is the wrong frame",[12,767,768],{},"Claude Design is not going to replace e-learning authoring tools. It is going to replace the parts of authoring tools that were template libraries dressed up as products — and, for a while, the parts of design tools that were template libraries dressed up as products. The parts that survive in e-learning — the parts worth paying for — are the parts that turn out to be e-learning-specific: instructional design opinions, learning objectives, retrieval practice, SCORM and xAPI, LMS integration, learner-state persistence, course-scale coherence, ongoing accessibility.",[12,770,771],{},"An authoring tool built around the same generation-plus-direct-editing idea Claude Design uses, but pointed at courses rather than prototypes — and bringing the L&D-specific plumbing — ends up doing what neither Claude Design alone nor a traditional template tool can do.",[12,773,774],{},"That is the bet LearnBuilder is making. Anthropic just made the case for it more clearly than I would have.",[211,776],{},[12,778,779],{},[16,780,781,782,785,786,728,791,543],{},"If you want to see what generation-inside-the-editor looks like in practice, the ",[171,783,223],{"href":220,"rel":784},[222]," includes the AI coding block in the interactive slideshow builder. Related: ",[171,787,790],{"href":788,"rel":789},"https:\u002F\u002Flearnbuilder.org\u002Fblog\u002Fai-coding-block-interactive-slides",[222],"Building an AI coding block into LearnBuilder",[171,792,795],{"href":793,"rel":794},"https:\u002F\u002Flearnbuilder.org\u002Fblog\u002Flearnbuilder-vs-vibe-coding",[222],"LearnBuilder vs vibe-coding your own learning content",{"title":226,"searchDepth":227,"depth":227,"links":797},[798,799,800,801,802,803,804],{"id":603,"depth":227,"text":604},{"id":619,"depth":227,"text":620},{"id":647,"depth":227,"text":648},{"id":698,"depth":227,"text":699},{"id":717,"depth":227,"text":718},{"id":751,"depth":227,"text":752},{"id":764,"depth":227,"text":765},"2026-04-26","Claude Design launched into Anthropic Labs in April with direct editing, design systems, and organization-scoped collaboration baked in. So what is left for e-learning authoring tools to do? An honest look at where Claude Design lands inside L&D — and where the gaps still are.","\u002Fblog\u002Fwill-claude-design-replace-elearning-authoring.webp",{},"\u002Fblog\u002Fwill-claude-design-replace-elearning-authoring",{"title":580,"description":806},"blog\u002Fwill-claude-design-replace-elearning-authoring",[247,813,250,574,572],"authoring","Dm06ycoLfadr9F6Qw7XzY6h_nvFVolB5GlqcawbqMKo",{"id":816,"title":817,"author":7,"body":818,"date":911,"description":912,"extension":239,"image":913,"meta":914,"navigation":242,"path":915,"seo":916,"stem":917,"tags":918,"__hash__":920},"blog\u002Fblog\u002Fai-coding-block-interactive-slides.md","Building an AI coding block into LearnBuilder: what works, what breaks, what's still open",{"type":9,"value":819,"toc":905},[820,823,826,829,833,836,839,842,845,849,852,858,864,870,874,877,883,889,892,896,899,902],[12,821,822],{},"After seeing a wave of Claude Design demos being used for e-learning, I started integrating an AI coding block into LearnBuilder. The idea is simple. You describe the interaction you want, the AI drafts the code, and the result drops into the interactive slide alongside the elements you built by hand.",[12,824,825],{},"The example I keep using: \"Create a simulation of a windmill to explain the correlation between wind speed and power output. Place a slider at the bottom of the page to let the learner explore the concept.\" Thirty seconds later, something shows up on screen.",[12,827,828],{},"And that is where the problems start.",[24,830,832],{"id":831},"the-iteration-problem","The iteration problem",[12,834,835],{},"The first output might look right, or it might not. The rotor rotates around the wrong origin. The animation is too slow, or spinning so fast it looks absurd. The blades are floating in the air, not attached to the tip of the pole. The slider works, but it controls the wrong variable.",[12,837,838],{},"So you nudge. \"Move the rotation pivot to the top of the pole.\" Sometimes you get a small correction. Often you get something entirely new. Different colors, different layout, sometimes a different kind of visualisation, because the AI rewrote the interaction from scratch instead of editing in place.",[12,840,841],{},"This is expensive. Complex interactions take real time to generate, and each nudge consumes tokens. For meaningful corrections you usually need to feed the original code back into the prompt so the AI can edit rather than reinvent. That means every iteration is a large call.",[12,843,844],{},"If the first draft is close, you are fine. If it is off in ways that need three or four rounds of nudging, you burn through significant tokens before you have something usable. And \"usable\" is the low bar. For a production course, you often need several more rounds after that.",[24,846,848],{"id":847},"what-already-works","What already works",[12,850,851],{},"A few things make this more practical than pure describe, generate, ship:",[12,853,854,857],{},[57,855,856],{},"AI-generated interactions live inside the slide's element tree."," The AI does not produce one opaque blob of code. It creates elements (shapes, animations, bindings) in the same structure you edit by hand. After generation you can select an element and change its color, reposition it, or resize it with the same controls you use for anything else. You are not locked out of manual edits.",[12,859,860,863],{},[57,861,862],{},"Element-level AI tweaks."," You can select a single element and ask the AI to modify just that. \"Make this rotor spin more slowly.\" \"Change the pivot of this shape.\" The AI gets scoped context, not the whole slide. This keeps tokens down and makes it much less likely that a small correction turns into a full rewrite, because the AI literally cannot touch the other elements.",[12,865,866,869],{},[57,867,868],{},"Custom JS interactions are exposed."," When the AI writes code for an interaction, that code is saved as a custom JS interaction in the slide. It is not hidden behind an opaque \"AI layer\". You (or a developer on your team) can open it, read it, edit it, extend it. If the AI gets you eighty percent of the way, you can cover the last twenty in code.",[24,871,873],{"id":872},"what-i-havent-solved-yet","What I haven't solved yet",[12,875,876],{},"Two hard problems are still open.",[12,878,879,882],{},[57,880,881],{},"Persistence."," If a learner interacts with the windmill simulation, moves the slider, sees the power curve, and then leaves the course, what happens when they come back? Do they see a fresh simulation, or the state they left? For a simple exploration tool, maybe it does not matter. For a scenario where the learner makes choices across multiple screens and expects to see those choices reflected on return, it matters a lot. AI-generated interactions do not get persistence for free. Every interaction needs a story for how its state is saved and restored, and that is not something the authoring prompt currently asks about.",[12,884,885,888],{},[57,886,887],{},"Reporting."," The adjacent problem. If the interaction is part of a graded exercise, the platform needs to know what the learner did. Did they land on the optimal wind speed? Did they explore the full range, or click once and leave? For a hand-built interaction, you wire this up deliberately. For an AI-generated one, there is no standard contract for what the interaction should report back to the LMS.",[12,890,891],{},"Both problems have the same underlying shape. AI generation is great for the visible surface of the interaction. E-learning needs the invisible plumbing too. Solving this probably means giving the AI a tighter spec to generate against, something like: here are the state variables you must expose, here are the events you must emit. That is the direction I am heading, but it is not there yet.",[24,893,895],{"id":894},"where-this-fits","Where this fits",[12,897,898],{},"An AI coding block is not a replacement for a designer who knows what good interactive learning looks like. It is a way to get from \"I want a windmill simulation\" to \"I have a windmill simulation I can edit\" in minutes instead of days. For simple interactions where the plumbing does not matter much, it is already useful. For graded, persistent, tracked interactions, there is real work left to do.",[12,900,901],{},"The honest position: this is a beta feature, and it will stay beta while the persistence and reporting story gets built out. I would rather ship something that works for exploratory interactions and is transparent about what it does not do yet, than ship a demo that looks magical and breaks the moment someone tries to grade against it.",[12,903,904],{},"If you want to try it, the AI coding block is live in LearnBuilder. I would especially love feedback from instructional designers who are already sketching the kind of interactions they would want to generate, because the next round of work is shaped by what people actually try to build with it.",{"title":226,"searchDepth":227,"depth":227,"links":906},[907,908,909,910],{"id":831,"depth":227,"text":832},{"id":847,"depth":227,"text":848},{"id":872,"depth":227,"text":873},{"id":894,"depth":227,"text":895},"2026-04-23","I built an AI coding block into LearnBuilder after seeing Claude Design demos used for e-learning. Here's what's working, what still breaks, and the hard problems I haven't solved yet: persistence, reporting, and the token cost of nudging.","\u002Fblog\u002Fai-coding-block-interactive-slides.webp",{},"\u002Fblog\u002Fai-coding-block-interactive-slides",{"title":817,"description":912},"blog\u002Fai-coding-block-interactive-slides",[247,574,813,572,919],"vibe coding","AQ5pON2GG2SGYZHawS6hHsvlGJL0jHVk-NSsH9mCoMc",{"id":922,"title":923,"author":7,"body":924,"date":1127,"description":1128,"extension":239,"image":1129,"meta":1130,"navigation":242,"path":173,"seo":1131,"stem":1132,"tags":1133,"__hash__":1136},"blog\u002Fblog\u002Fai-assisted-hand-built-or-both-learnbuilder.md","AI-assisted, hand-built, or both: how to choose inside LearnBuilder",{"type":9,"value":925,"toc":1118},[926,929,932,936,939,942,945,959,962,966,969,972,975,981,985,988,991,994,997,1001,1004,1007,1010,1021,1024,1028,1031,1034,1037,1041,1044,1049,1062,1067,1081,1086,1097,1101,1104,1107,1110,1113],[12,927,928],{},"Most authoring tools force a binary choice: either you build everything by hand, or you hand everything to AI and hope for the best. LearnBuilder doesn't work that way. You can generate a full course from a document in minutes, you can build every screen manually with full control, or you can do something in between — AI drafts the structure, you refine the pedagogy.",[12,930,931],{},"The question isn't whether AI is \"good\" or \"bad\" for course authoring. The question is: which workflow fits your content, your timeline, your regulatory constraints, and your comfort level with what you're uploading?",[24,933,935],{"id":934},"when-full-ai-generation-makes-sense","When full AI generation makes sense",[12,937,938],{},"If you have a 40-page policy document, a technical specification, or a product manual that needs to become a compliance course by Friday, AI generation is not a shortcut — it's the only realistic option. Upload the document, set a few parameters (course length, question density, tone), and LearnBuilder generates a complete course structure with explanatory content, retrieval questions, and knowledge checks.",[12,940,941],{},"The time savings are real. A course that would take two days to outline, draft, and structure manually can be generated in under ten minutes. Not \"generated and then completely rewritten\" — generated to a standard where the instructional designer's job shifts from creation to refinement. You're editing, not authoring from scratch.",[12,943,944],{},"This works best when:",[455,946,947,950,953,956],{},[458,948,949],{},"The source material is factual and well-structured",[458,951,952],{},"The content doesn't require nuanced judgment calls (e.g., compliance training, technical onboarding, process documentation)",[458,954,955],{},"You need a first draft fast and can refine iteratively",[458,957,958],{},"The audience is internal and the stakes for perfection are moderate",[12,960,961],{},"It works less well when the source material is ambiguous, when the pedagogy requires careful sequencing that AI might not infer correctly, or when the content involves sensitive interpersonal scenarios that need a human touch. AI can generate a harassment prevention course, but whether it should is a different question.",[24,963,965],{"id":964},"the-gdpr-question-what-are-you-actually-uploading","The GDPR question: what are you actually uploading?",[12,967,968],{},"Here's the part most vendors skip: if you're uploading documents to generate courses, you're sending that content to an AI model. If that document contains personal data, trade secrets, or anything covered by GDPR, HIPAA, or your organisation's data governance policies, you need to know where it's going and what happens to it.",[12,970,971],{},"LearnBuilder uses European AI infrastructure with full GDPR compliance. That means your documents are processed on servers in the EU, they're not used to train models, and they're deleted after processing. For L&D teams in regulated industries — healthcare, finance, government — this isn't a nice-to-have. It's the difference between \"we can use this tool\" and \"legal won't approve it.\"",[12,973,974],{},"Compare this to tools that route content through US-based APIs with vague data retention policies, or worse, tools that explicitly state uploaded content may be used for model improvement. If you're in pharmaceuticals and you upload a document about an unreleased drug, or you're in finance and you upload internal risk assessment guidelines, you need to know that content isn't being stored, analysed, or fed back into a training corpus.",[12,976,977,980],{},[57,978,979],{},"The honest limitation:"," even with GDPR-compliant infrastructure, uploading sensitive documents to any AI system introduces risk. If your content is classified, if it contains patient data, if it's subject to export controls — don't upload it to an AI tool, full stop. Use manual authoring. The compliance risk isn't worth the time savings.",[24,982,984],{"id":983},"when-manual-authoring-is-the-right-choice","When manual authoring is the right choice",[12,986,987],{},"Some courses shouldn't be AI-generated, even when the technology is good enough. Soft skills training, leadership development, scenario-based learning — these benefit from the deliberate pacing, the careful choice of examples, and the pedagogical decisions that an experienced instructional designer makes instinctively but that AI infers statistically.",[12,989,990],{},"LearnBuilder's manual authoring mode gives you the same tools as AI generation — branching scenarios, embedded questions, rich media support — but you're building each screen yourself. You're not fighting an AI-generated structure you didn't want. You're not editing around awkward phrasing that sounds almost right but isn't quite.",[12,992,993],{},"This is also the right choice when you're working with content that doesn't exist in document form. If the course is based on interviews, if it's synthesising knowledge from multiple stakeholders, if it's teaching a skill that's mostly tacit — you need to author manually because there's no source document to upload.",[12,995,996],{},"Manual authoring is slower. A course that AI could draft in ten minutes might take a day to build by hand. But \"slower\" doesn't mean \"worse.\" It means you're making deliberate choices about structure, sequence, and emphasis that AI would make probabilistically. For high-stakes content, that's worth the time.",[24,998,1000],{"id":999},"the-hybrid-workflow-ai-drafts-humans-refine","The hybrid workflow: AI drafts, humans refine",[12,1002,1003],{},"The most common workflow in practice is neither pure AI nor pure manual — it's AI-generated structure with human refinement. Upload a document, generate the course, then go through and adjust the pedagogy, rewrite the questions, add examples, reorder sections, and insert media.",[12,1005,1006],{},"This is faster than manual authoring and better than unedited AI output. You're not starting from a blank screen, but you're also not shipping the first draft. The AI handles the grunt work — extracting key points, generating initial questions, structuring the content — and you handle the judgment calls.",[12,1008,1009],{},"In practice, this looks like:",[455,1011,1012,1015,1018],{},[458,1013,1014],{},"Generating a course from a technical document, then rewriting the questions to be more scenario-based",[458,1016,1017],{},"Using AI to create the core content, then manually adding branching scenarios for decision-making practice",[458,1019,1020],{},"Letting AI draft the explanatory text, then editing for tone and adding real-world examples from your organisation",[12,1022,1023],{},"The time savings here are still significant — maybe 50-60% faster than pure manual authoring — but the output quality is higher than pure AI generation because a human is making the pedagogical decisions.",[24,1025,1027],{"id":1026},"what-about-uploading-documents-you-didnt-write","What about uploading documents you didn't write?",[12,1029,1030],{},"One underrated use case: generating courses from third-party content that's already public or licensed. If you're training employees on a software tool and the vendor has published a user guide, you can upload that guide and generate a course. If you're onboarding new hires on industry regulations and there's a publicly available compliance framework, you can use that as source material.",[12,1032,1033],{},"This sidesteps the data governance question — you're not uploading internal documents, you're using content that's already in the public domain or that you have explicit permission to use. The AI generation workflow becomes a way to transform reference material into structured learning without the compliance risk.",[12,1035,1036],{},"The caveat: you still need to verify accuracy. Just because a document is public doesn't mean it's correct, and just because AI generated a course from it doesn't mean the course is pedagogically sound. You're still responsible for the output.",[24,1038,1040],{"id":1039},"how-to-decide-for-your-next-course","How to decide for your next course",[12,1042,1043],{},"Here's a decision framework:",[12,1045,1046],{},[57,1047,1048],{},"Use AI generation when:",[455,1050,1051,1054,1057,1059],{},[458,1052,1053],{},"You have a well-structured source document",[458,1055,1056],{},"The content is factual and the pedagogy is straightforward",[458,1058,955],{},[458,1060,1061],{},"The document doesn't contain sensitive data, or you've confirmed GDPR compliance is sufficient for your governance requirements",[12,1063,1064],{},[57,1065,1066],{},"Use manual authoring when:",[455,1068,1069,1072,1075,1078],{},[458,1070,1071],{},"The content requires careful pedagogical sequencing",[458,1073,1074],{},"You're teaching soft skills, leadership, or scenario-based decision-making",[458,1076,1077],{},"The content doesn't exist in document form",[458,1079,1080],{},"You're working with classified or highly sensitive material that shouldn't be uploaded anywhere",[12,1082,1083],{},[57,1084,1085],{},"Use hybrid workflows when:",[455,1087,1088,1091,1094],{},[458,1089,1090],{},"You want the speed of AI generation but need human refinement for quality",[458,1092,1093],{},"The source material is good but needs pedagogical improvement",[458,1095,1096],{},"You're comfortable uploading the document but not comfortable shipping unedited AI output",[24,1098,1100],{"id":1099},"the-thing-nobody-says-about-ai-authoring-tools","The thing nobody says about AI authoring tools",[12,1102,1103],{},"Most AI authoring tools are optimised for the demo, not the workflow. They show you a polished course generated in seconds, but they don't show you the instructional designer spending an hour editing the output to make it usable. They don't talk about what happens when you upload a poorly structured document and the AI generates a poorly structured course. They don't address the data governance question because it's inconvenient.",[12,1105,1106],{},"LearnBuilder's approach is: AI is a drafting tool, not a replacement for instructional design. It's faster than starting from scratch, but it's not a magic button that produces perfect courses. You still need to make pedagogical decisions. You still need to verify accuracy. You still need to think about whether the content you're uploading should be uploaded at all.",[12,1108,1109],{},"The benefit of hosting on European AI infrastructure with GDPR compliance isn't that it makes every document safe to upload — it's that it makes the decision framework clearer. If the document is internal but not classified, if it's covered by GDPR but not by stricter regulations, if it's something you'd be comfortable storing on a European server — then AI generation is an option. If any of those conditions don't hold, use manual authoring.",[12,1111,1112],{},"The honest answer is: most L&D teams will end up using all three workflows depending on the course. AI generation for compliance training and technical onboarding. Manual authoring for leadership development and soft skills. Hybrid workflows for everything in between.",[12,1114,1115],{},[16,1116,1117],{},"Try LearnBuilder free for 14 days and see which workflow fits your content. No credit card required.",{"title":226,"searchDepth":227,"depth":227,"links":1119},[1120,1121,1122,1123,1124,1125,1126],{"id":934,"depth":227,"text":935},{"id":964,"depth":227,"text":965},{"id":983,"depth":227,"text":984},{"id":999,"depth":227,"text":1000},{"id":1026,"depth":227,"text":1027},{"id":1039,"depth":227,"text":1040},{"id":1099,"depth":227,"text":1100},"2026-04-19","LearnBuilder supports AI course generation, full manual authoring, and hybrid workflows where AI drafts and humans refine. Here's how to choose the right approach for your content, your constraints, and your comfort level with uploading documents to AI.","\u002Fblog\u002Fai-assisted-hand-built-or-both-learnbuilder.webp",{},{"title":923,"description":1128},"blog\u002Fai-assisted-hand-built-or-both-learnbuilder",[247,251,1134,250,1135],"GDPR","workflows","_YbKjjba1i3GbjdBfc-E5SWFfuXHy599GdRKXyGE8TQ",{"id":1138,"title":1139,"author":7,"body":1140,"date":1285,"description":1286,"extension":239,"image":1287,"meta":1288,"navigation":242,"path":1289,"seo":1290,"stem":1291,"tags":1292,"__hash__":1294},"blog\u002Fblog\u002Fcompliance-training-development-time.md","Compliance training doesn't have to be boring — but it often should be quick",{"type":9,"value":1141,"toc":1278},[1142,1145,1148,1151,1155,1166,1169,1172,1175,1178,1182,1185,1188,1191,1194,1197,1200,1204,1207,1210,1213,1216,1219,1222,1226,1229,1235,1241,1247,1250,1254,1257,1260,1263,1266,1268],[12,1143,1144],{},"There is a debate that surfaces regularly in instructional design circles about compliance training. One side says it is fundamentally a checkbox exercise — employees know it, managers know it, legal knows it — and the right response is to make it as frictionless as possible. Click through, pass the quiz, done. The other side argues that compliance training covers genuinely important topics and deserves proper instructional design: scenarios, practice, meaningful feedback.",[12,1146,1147],{},"Both positions are right about something and wrong about something, and the disagreement usually stems from conflating very different types of compliance content into a single category.",[12,1149,1150],{},"This article is about what is actually achievable — from the author's side and the learner's side — and where the development effort is genuinely worth spending.",[24,1152,1154],{"id":1153},"not-all-compliance-topics-are-the-same","Not all compliance topics are the same",[12,1156,1157,1158,1161,1162,1165],{},"The most useful distinction I have come across is not between \"engaging\" and \"boring\" compliance training. It is between compliance topics where the training is primarily about ",[16,1159,1160],{},"demonstrating"," that staff have been informed, and topics where the training is primarily about ",[16,1163,1164],{},"changing how staff behave"," in specific situations.",[12,1167,1168],{},"Fire safety induction. Annual data protection refreshers. General health and safety acknowledgements. For most organisations, for most employees, these fall into the first category. The regulatory requirement is that staff have received the information. Whether the e-learning module was well-designed has essentially no bearing on whether staff evacuate correctly — that depends on practice drills, physical signage, and muscle memory, none of which a click-through module produces.",[12,1170,1171],{},"Trying to make fire safety induction deeply engaging is not a design problem. It is a category error.",[12,1173,1174],{},"But consider cybersecurity awareness training, where the actual threat is a staff member clicking a convincing phishing link under time pressure. Or safeguarding training, where staff need to recognise warning signs in ambiguous situations and know what to do with that recognition. Or conflict of interest training in a financial services context, where the decisions are genuinely difficult and the consequences of getting them wrong are serious. These topics involve real judgment calls in real situations, and a module that delivers information passively and asks whether you understood it is not doing the job.",[12,1176,1177],{},"The design question is not \"how do I make this compliance module more engaging?\" It is \"does this topic require behavior change, and if so, what kind of practice does that require?\"",[24,1179,1181],{"id":1180},"the-authors-problem-time-versus-ambition","The author's problem: time versus ambition",[12,1183,1184],{},"The reason most compliance training ends up as click-through content is not that instructional designers lack the skills or the intentions to do better. It is that genuinely interactive content takes time to build, and compliance training timelines are almost never generous.",[12,1186,1187],{},"A well-constructed branching scenario in a traditional authoring tool — one with enough decision points to give learners meaningful practice, realistic characters, consequence paths that illuminate the learning objective — might take a day or two for a single scenario of moderate complexity. Multiply that by a course with five or six modules, and you are looking at a production schedule that most compliance training projects simply do not have.",[12,1189,1190],{},"The result is a rational but unfortunate trade-off: designers produce what is achievable in the time available, which usually means text-heavy content and a multiple-choice quiz at the end. The quiz is graded at 80% because someone decided 80% was the right number. Nobody is sure whether any of it changes anything.",[12,1192,1193],{},"This is where the development time question becomes interesting, because the bottleneck has shifted.",[12,1195,1196],{},"AI-assisted authoring does not eliminate design judgment, but it substantially changes the time cost of production. A dialogue scenario that would have taken a day to script, branch, and build can be set up in minutes — you describe the situation, define the characters and how they should behave, specify what the learner should demonstrate, and the AI handles the conversation dynamically. You do not write every possible response path. You write a strong scenario and clear learning outcomes.",[12,1198,1199],{},"That changes the calculation. If the production barrier to a realistic conversation scenario is a day's work, it gets cut from the compliance module when the schedule tightens. If the barrier is twenty minutes, the decision looks different.",[24,1201,1203],{"id":1202},"what-more-meaningful-actually-means-for-the-learner","What \"more meaningful\" actually means for the learner",[12,1205,1206],{},"It is worth being honest about what learner engagement in compliance training can realistically achieve — because the goal is not engagement for its own sake.",[12,1208,1209],{},"For the click-through-and-pass category of compliance content, the learner experience is largely irrelevant to the outcome. Staff will complete it because they have to. They will retain approximately as much as they retain from any passive reading experience, which is not much. That is fine, because the goal was not retention — it was a completion record.",[12,1211,1212],{},"For the behavior-change category, what matters is not whether the module is visually polished or gamified or fun. What matters is whether learners encounter realistic situations that require them to make a decision, get feedback on that decision, and understand why the correct approach is what it is. That is a very specific type of interactivity, and a lot of what gets labelled \"engaging compliance training\" does not provide it.",[12,1214,1215],{},"A scenario that puts a learner in a realistic situation — a message that might be a phishing attempt, a colleague asking them to bend a procedure, a client disclosure that might cross a regulatory line — and asks them to respond, then shows them the consequence of that response, is doing something genuinely useful. The learner is practising the judgment, not reading about it.",[12,1217,1218],{},"An animated infographic with knowledge-check questions at the end is more visually appealing than a text block, but it is not doing something fundamentally different from a pedagogical standpoint.",[12,1220,1221],{},"The distinction matters because it prevents conflating production quality with instructional quality. A well-designed scenario built with modest visuals outperforms a beautifully produced module that delivers content passively.",[24,1223,1225],{"id":1224},"what-is-actually-achievable-with-less-time","What is actually achievable with less time",[12,1227,1228],{},"A realistic picture for an organisation building compliance training in LearnBuilder looks something like this.",[12,1230,1231,1234],{},[57,1232,1233],{},"For information-delivery topics",", the right investment is clarity and brevity. Clean, well-structured content, broken into short lessons, with retrieval practice questions interspersed throughout rather than saved for a final quiz. That last point is small but meaningful — questions mid-lesson, before the answer has been restated, strengthen retention measurably compared to end-of-course assessments. It takes no extra production time; it just requires positioning questions differently.",[12,1236,1237,1240],{},[57,1238,1239],{},"For judgment-based topics",", a single well-designed dialogue scenario per module is often sufficient to deliver genuine practice. Set up the situation, introduce the characters, define the learning outcomes, let the AI handle the conversation dynamically. Learners respond in their own words, the scenario responds accordingly, and the Assessment Agent evaluates the conversation against your criteria at the end. One scenario, set up in twenty minutes, gives learners more meaningful practice than a branching flowchart that took a day to build and still only covers the paths the author anticipated.",[12,1242,1243,1246],{},[57,1244,1245],{},"For the quiz itself",", open-ended short-answer questions with AI grading give better signal than multiple choice. A learner who can select the correct answer from four options does not necessarily understand the principle. A learner who can articulate why a particular action was correct or incorrect, in their own words, probably does. The Assessment Agent grades these responses and provides individual written feedback, so the grading overhead that would make this impractical at scale is handled.",[12,1248,1249],{},"None of this requires a large production budget or a long timeline. It requires knowing which topics warrant which approach, and a tool that makes the right approach accessible within realistic time constraints.",[24,1251,1253],{"id":1252},"the-honest-summary","The honest summary",[12,1255,1256],{},"Compliance training is not going to become anyone's favourite professional development experience, and that is not a failure of instructional design. For a significant proportion of compliance content, the goal is a completion record, and optimising for that is a legitimate choice.",[12,1258,1259],{},"But there is a category of compliance topics — the ones where staff genuinely need to make better decisions in difficult situations — where the difference between click-through content and meaningful practice is measurable. Not in engagement scores. In whether staff do the right thing when a real situation arises.",[12,1261,1262],{},"The argument for investing in those topics is not that learning should be fun. It is that the training exists to change something, and passive content does not change behavior.",[12,1264,1265],{},"If the barrier has historically been production time, that barrier is lower than it was. Whether to invest the time in better design is still a judgment call — but it should be made deliberately, based on what the training is actually for.",[211,1267],{},[12,1269,1270],{},[16,1271,1272,1273,1277],{},"LearnBuilder is free to try at ",[171,1274,1276],{"href":220,"rel":1275},[222],"learnbuilder.org",". No credit card required.",{"title":226,"searchDepth":227,"depth":227,"links":1279},[1280,1281,1282,1283,1284],{"id":1153,"depth":227,"text":1154},{"id":1180,"depth":227,"text":1181},{"id":1202,"depth":227,"text":1203},{"id":1224,"depth":227,"text":1225},{"id":1252,"depth":227,"text":1253},"2026-04-16","The compliance training debate tends to collapse into two camps: people who think it should all be click-through checkboxes, and people who think every module deserves scenario-based design. Both are wrong in interesting ways.","\u002Fblog\u002Fcompliance-training-development-time.webp",{},"\u002Fblog\u002Fcompliance-training-development-time",{"title":1139,"description":1286},"blog\u002Fcompliance-training-development-time",[250,1293],"compliance","-SwxVv15icroxlCwUJaNS2RGgEsqh9Ac_G0uOjjwCjg",{"id":1296,"title":1297,"author":7,"body":1298,"date":1446,"description":1447,"extension":239,"image":1448,"meta":1449,"navigation":242,"path":1450,"seo":1451,"stem":1452,"tags":1453,"__hash__":1455},"blog\u002Fblog\u002Flearnbuilder-vs-vibe-coding.md","LearnBuilder vs. vibe-coding your own learning content",{"type":9,"value":1299,"toc":1436},[1300,1303,1306,1309,1313,1316,1319,1322,1326,1329,1332,1335,1338,1342,1345,1348,1351,1355,1358,1361,1364,1368,1371,1374,1377,1380,1384,1387,1390,1393,1397,1400,1403,1406,1409,1412,1416,1419,1422,1425,1428,1430],[12,1301,1302],{},"Something has shifted in the last year or two. If you can describe what you want in plain language, you can get working code — a branching scenario, a drag-and-drop activity, an interactive quiz — without writing a line yourself. The people building learning content have noticed. The question that keeps coming up, especially from technically curious instructional designers and developers who have drifted into L&D, is a reasonable one: why use a purpose-built authoring tool at all?",[12,1304,1305],{},"Vibe-coding your own learning content is a real option now. It is worth taking seriously rather than dismissing. But it comes with a set of trade-offs that are worth understanding before you commit to an approach — and some of them only become visible after you are three modules into a course.",[12,1307,1308],{},"This is an honest comparison. I built LearnBuilder, so I have a position, and I will be upfront about where the vibe-coding approach genuinely wins.",[24,1310,1312],{"id":1311},"what-vibe-coding-your-learning-content-actually-means","What vibe-coding your learning content actually means",[12,1314,1315],{},"\"Vibe-coding\" in this context means using an AI coding assistant — Claude, Cursor, ChatGPT, v0, or similar — to generate the HTML, CSS, and JavaScript for your learning activities by describing what you want. You might build individual interactive exercises as standalone files, embed them in a course platform, or string them together into something that looks like a course.",[12,1317,1318],{},"At its best, this produces genuinely impressive results. An interactive simulation of a software interface. A branching scenario rendered exactly the way you want it. A custom calculator or data visualisation embedded in a lesson. The output can be polished, specific, and hard to replicate quickly in a template-driven tool.",[12,1320,1321],{},"So what is the case for a purpose-built authoring tool?",[24,1323,1325],{"id":1324},"learning-design-is-a-different-problem-from-ui-design","Learning design is a different problem from UI design",[12,1327,1328],{},"The first gap is easy to miss until you are already building. Vibe-coding tools are good at generating interfaces. They are not opinionated about instructional design — they have no model of what makes a learning experience effective, what the learning outcomes are, or how the content should be sequenced to build toward them.",[12,1330,1331],{},"When you use a tool like LearnBuilder to generate content, the AI is working from instructional principles — it understands the difference between a knowledge check and a performance-based activity, enforces content variety so learners are not reading text blocks for twenty minutes in a row, and generates retrieval practice and scenario-based exercises alongside explanatory content. The lesson structure is driven by learning objectives, not by what is easy to render.",[12,1333,1334],{},"With a vibe-coding approach, that work is entirely on you. The AI will generate whatever you describe — but describing a well-structured learning experience requires knowing what a well-structured learning experience looks like. That is the designer's job, and the tool does not help you do it.",[12,1336,1337],{},"For experienced instructional designers, this is not necessarily a problem — they bring that knowledge themselves. For everyone else, it is a significant gap.",[24,1339,1341],{"id":1340},"consistency-across-a-course-is-harder-than-it-looks","Consistency across a course is harder than it looks",[12,1343,1344],{},"A single interactive exercise is relatively easy to produce with an AI coding assistant. A full course — ten or fifteen lessons, each with explanatory content, knowledge checks, scenario-based activities, and a consistent visual identity — is a different problem.",[12,1346,1347],{},"Each generation session starts more or less fresh. Getting consistent typography, spacing, color, component styles, and interaction patterns across a set of files generated across multiple sessions requires careful prompting, shared CSS, and manual review. Even with a well-maintained design system, drift accumulates. A button that looks slightly different. A feedback message styled differently. A layout that doesn't quite match.",[12,1349,1350],{},"In LearnBuilder, consistency is handled structurally. Your brand colors, fonts, and component styles are configured once at the account level and applied across every lesson. If you change the primary color, it updates everywhere. The course looks like a course — not like a collection of individually generated exercises that happen to be on the same topic.",[24,1352,1354],{"id":1353},"accessibility-requires-sustained-attention-not-just-initial-generation","Accessibility requires sustained attention, not just initial generation",[12,1356,1357],{},"Generated code does not reliably produce accessible content. AI coding assistants will produce accessible markup when explicitly prompted for it, but accessibility in learning content is not a one-time decision — it is an ongoing property that has to be maintained across every element you add, every interaction you wire up, every video you embed.",[12,1359,1360],{},"Missing alt text on an image. A drag-and-drop interaction with no keyboard equivalent. A video without captions. A color contrast ratio that fails WCAG AA. These are easy to introduce and easy to miss when you are focused on the interaction design rather than the accessibility audit.",[12,1362,1363],{},"LearnBuilder includes a built-in accessibility checker that flags these issues before you publish — missing captions, missing alt text, hover-only interactions that are not keyboard accessible, low contrast. It is not a substitute for a proper accessibility review, but it catches the common failures that generated content tends to introduce.",[24,1365,1367],{"id":1366},"teams-and-version-control","Teams and version control",[12,1369,1370],{},"If you are working alone, vibe-coded content lives wherever you keep files. If you are working with others, it gets complicated quickly.",[12,1372,1373],{},"Sharing generated HTML files via email or a shared drive works until someone edits the wrong version. A Git workflow solves that, but most L&D teams do not work in Git, and most instructional designers do not want to. Even on teams that do, merge conflicts in generated HTML and CSS are unpleasant to resolve.",[12,1375,1376],{},"LearnBuilder has real-time multi-author collaboration built in. Multiple people can work on different lessons in a course simultaneously. Changes are saved automatically and visible immediately. There is no file-sharing step, no version confusion, and no setup required — it works the way collaborative writing tools work, not the way a software project works.",[12,1378,1379],{},"For solo designers, this does not matter. For any team of two or more, it matters a great deal.",[24,1381,1383],{"id":1382},"reusable-exercises-and-activities","Reusable exercises and activities",[12,1385,1386],{},"Learning content has patterns that repeat. An activity format that works for one lesson — a drag-and-drop matching exercise, a multi-step scenario, a fill-in-the-blank vocabulary check — is worth reusing rather than regenerating from scratch each time.",[12,1388,1389],{},"With a vibe-coding approach, reuse means copying and editing code. That works, but it creates duplicate files that diverge over time. If you find a bug or want to improve the format, you fix it in one file and then have to remember to fix it everywhere else.",[12,1391,1392],{},"LearnBuilder structures content as blocks within lessons. A quiz format you have configured, a dialogue scenario structure, an interactive slide template — these can be reused across lessons and courses without duplication. Updates to a block type propagate rather than requiring manual edits across multiple files.",[24,1394,1396],{"id":1395},"simple-changes-in-a-visual-tool","Simple changes in a visual tool",[12,1398,1399],{},"This one genuinely cuts in favor of the vibe-coding approach, and it is worth acknowledging.",[12,1401,1402],{},"There are changes that are faster in a visual editing environment — moving a block, resizing an image, tweaking a heading — than they are in a code editor, even with AI assistance. The round-trip of describing a change, waiting for regenerated code, and checking the result is slower than dragging an element to a new position. When a stakeholder wants a quick layout change or a content edit before a review session, a visual tool is simply faster for that kind of work.",[12,1404,1405],{},"Where vibe-coding wins is at the edges of what a standard authoring tool can do. If you want an interaction type that does not exist as a built-in block, generated code is the fastest path to something that works.",[12,1407,1408],{},"This is why LearnBuilder includes a Custom Embed block — HTML, CSS, and JavaScript, running in a sandboxed iframe inside the lesson. You can use an AI coding assistant to generate whatever you want, paste it in, and have it run inside a properly structured course with all the surrounding infrastructure — consistent styling, progress tracking, completion reporting — handled by the platform. AI generation is available directly in the block, so you can describe what you want and iterate from there without leaving the editor.",[12,1410,1411],{},"The practical result is that you do not have to choose between visual authoring efficiency and code-level flexibility. The Custom Embed block is the escape hatch for the cases where a standard block type is not enough.",[24,1413,1415],{"id":1414},"where-vibe-coding-genuinely-wins","Where vibe-coding genuinely wins",[12,1417,1418],{},"The honest answer is that for a technically capable individual designer building something that needs to look and behave in a very specific way — and who is not concerned with team collaboration, long-term maintenance, or instructional scaffolding — vibe-coding is a real option.",[12,1420,1421],{},"It gives you complete control over the output. There are no template constraints, no platform decisions about what is or is not a supported interaction type, no pricing tier that limits how many courses you can have. If you are comfortable in a code editor and you know your instructional design, the ceiling is higher than any authoring tool.",[12,1423,1424],{},"The cost of that ceiling is everything that purpose-built tools handle in the background: the design system consistency, the accessibility checking, the team collaboration, the LMS integration, the learner tracking, the enrollment and completion management. Each of those is a problem you own when you own the code.",[12,1426,1427],{},"For most teams building learning content at any kind of scale, that trade-off does not make sense. For a technically curious solo designer building a specific, high-craft experience, it might.",[211,1429],{},[12,1431,1432,1433,224],{},"If you are evaluating authoring tools and want to see what the Custom Embed block looks like in practice alongside the rest of LearnBuilder's capabilities, the ",[171,1434,223],{"href":220,"rel":1435},[222],{"title":226,"searchDepth":227,"depth":227,"links":1437},[1438,1439,1440,1441,1442,1443,1444,1445],{"id":1311,"depth":227,"text":1312},{"id":1324,"depth":227,"text":1325},{"id":1340,"depth":227,"text":1341},{"id":1353,"depth":227,"text":1354},{"id":1366,"depth":227,"text":1367},{"id":1382,"depth":227,"text":1383},{"id":1395,"depth":227,"text":1396},{"id":1414,"depth":227,"text":1415},"2026-04-15","AI has made it genuinely possible to build interactive learning experiences by describing what you want in plain language. So why use a purpose-built authoring tool at all? The answer is more nuanced than it might seem.","\u002Fblog\u002Flearnbuilder-vs-vibe-coding.webp",{},"\u002Fblog\u002Flearnbuilder-vs-vibe-coding",{"title":1297,"description":1447},"blog\u002Flearnbuilder-vs-vibe-coding",[250,1454],"tools","5An1ibNZZXJObDZ8E2DnGxUm3a_9xAGR3B7NmFIdufU",{"id":1457,"title":1458,"author":7,"body":1459,"date":1743,"description":1744,"extension":239,"image":1745,"meta":1746,"navigation":242,"path":1747,"seo":1748,"stem":1749,"tags":1750,"__hash__":1751},"blog\u002Fblog\u002Flearnbuilder-for-instructional-designers.md","LearnBuilder for instructional designers",{"type":9,"value":1460,"toc":1732},[1461,1470,1473,1476,1480,1483,1486,1497,1500,1506,1512,1518,1524,1527,1531,1534,1537,1552,1555,1558,1572,1578,1584,1597,1611,1615,1618,1621,1627,1633,1636,1640,1643,1646,1649,1653,1659,1662,1665,1669,1672,1675,1679,1682,1685,1689,1692,1695,1699,1702,1705,1711,1717,1723,1725],[12,1462,1463,1464,1469],{},"Most of what gets written about AI authoring tools is aimed at people who have never built a course before. The promise is always some version of the same thing: paste in a document, click generate, get a course. That capability is real and useful — we wrote about it ",[171,1465,1468],{"href":1466,"rel":1467},"https:\u002F\u002Flearnbuilder.org\u002Fblog\u002Fhow-to-turn-a-document-into-staff-training",[222],"here"," — but it is not the most interesting thing about LearnBuilder for someone who already knows instructional design.",[12,1471,1472],{},"This post is for the people who know what a branching scenario is and why it matters. Who have spent time in Storyline wiring up triggers. Who understand the difference between a knowledge check and a performance-based assessment. And who are wondering whether LearnBuilder has enough depth to be worth adopting — or whether it is just another tool built for beginners that you will outgrow in a month.",[12,1474,1475],{},"The short answer is that LearnBuilder was designed with this question in mind. Here is what the authoring environment actually gives you.",[24,1477,1479],{"id":1478},"ai-dialogue-scenario-based-learning-without-the-branching-overhead","AI Dialogue: scenario-based learning without the branching overhead",[12,1481,1482],{},"Scenario-based learning is one of the best-evidenced approaches in instructional design. It is also one of the most time-consuming things to build well. A realistic branching scenario in a traditional authoring tool involves writing multiple response paths, anticipating learner choices, creating consequence slides for each branch, and testing the whole thing for logic errors. For a single scenario of moderate complexity, that can easily be a day's work.",[12,1484,1485],{},"LearnBuilder's AI Dialogue block approaches this differently. Instead of pre-authored branches, the learner interacts in free-form text with AI-played characters. You write the scenario context and define the characters — their name, role, and how they should behave — and the AI plays them dynamically throughout the conversation. The learner can say anything. The scenario responds accordingly.",[12,1487,1488,1489,1492,1493,1496],{},"This shifts the design effort from ",[16,1490,1491],{},"writing every possible path"," to ",[16,1494,1495],{},"writing a strong scenario and clear learning outcomes",". The learning outcomes you define are what the Assessment Agent evaluates against when the dialogue ends — it scores the conversation and gives the learner written feedback. As the designer, you set the passing threshold.",[12,1498,1499],{},"A few things worth knowing for design purposes:",[12,1501,1502,1505],{},[57,1503,1504],{},"Character poses."," Characters can be rendered standing (as a cutout against a background) or sitting, where the system composites them naturally into a desk or workspace setting. Individual emotion variants can be regenerated if they don't look right, without affecting the rest of the character.",[12,1507,1508,1511],{},[57,1509,1510],{},"Observer mode."," You can set a dialogue to observer mode, where the learner watches two AI-played characters have a conversation rather than participating. This is useful when you want to model a scenario — show what good looks like, or deliberately show a failure — before asking learners to practice themselves.",[12,1513,1514,1517],{},[57,1515,1516],{},"Linked dialogues."," Dialogues can be chained into multi-part sequences. The context carries over from one to the next, so you can design a scenario that unfolds in stages.",[12,1519,1520,1523],{},[57,1521,1522],{},"Audio input."," Learners can speak their responses using a microphone instead of typing. This is particularly well-suited to language training, presentation coaching, or any scenario where verbal communication is specifically what you are trying to develop.",[12,1525,1526],{},"The honest limitation: the AI Dialogue block is powerful for naturalistic conversation practice, but it is not the right tool for every scenario. If your use case requires very tightly controlled decision points — where the specific choice the learner makes should determine a specific consequence — the Interactive Slides block (below) gives you that precision.",[24,1528,1530],{"id":1529},"interactive-slides-a-canvas-editor-with-a-full-trigger-action-system","Interactive Slides: a canvas editor with a full trigger-action system",[12,1532,1533],{},"If you have used Articulate Storyline, the Interactive Slides block will feel immediately familiar. It is a visual canvas editor where you place and position elements freely, then wire up behaviors using a trigger-action system.",[12,1535,1536],{},"The trigger list covers what you would expect — click, hover, hover out, double-click — plus video events (play, pause, ended, specific timestamp), drag-and-drop events (drag start, drop, correct drop, incorrect drop), timer completion, scroll position, and animation completion. Actions include navigation between slides, show\u002Fhide\u002Ftoggle on individual elements or element groups via tags, text input validation, variable manipulation, timer control, and video playback control.",[12,1538,1539,1540,1543,1544,1547,1548,1551],{},"Variables are a first-class element type. You can place a variable on the canvas and wire up actions to increment, decrement, toggle, or reset it — which means you can build scored activities, track multi-step choices, and display results on a dedicated slide. The scoring system has dedicated ",[665,1541,1542],{},"score",", ",[665,1545,1546],{},"maxScore",", and ",[665,1549,1550],{},"passingScore"," variables built in.",[12,1553,1554],{},"The element library covers text, images, video, audio, shapes, icons, buttons, text inputs, textareas, checkboxes, drag items, drop zones, timers, variables, scroll containers, and Lottie animations. Scroll containers are worth noting specifically: they let you create a scrollable region inside a fixed-size slide — which is how you would build a simulated email inbox, a document review interface, or any other content that needs to scroll within a defined space.",[12,1556,1557],{},"A few interaction design details that experienced designers will care about:",[12,1559,1560,1563,1564,1567,1568,1571],{},[57,1561,1562],{},"Element tags."," You can tag multiple elements with a shared label and then target the tag in actions. A single \"hide\" action can hide every element tagged ",[665,1565,1566],{},"feedback-correct"," or ",[665,1569,1570],{},"option"," at once, rather than wiring the same action to each element individually.",[12,1573,1574,1577],{},[57,1575,1576],{},"Element animations."," Independent of trigger interactions, elements can have looping animations: pulse, bounce, float, shake, wiggle, spin, fade, glow. These are useful for drawing attention to an element or giving a simulation life without requiring a trigger.",[12,1579,1580,1583],{},[57,1581,1582],{},"Multi-select editing."," Shift-click or drag-select to select multiple elements and change shared properties (fill, stroke, text color, font size) across all of them at once. Useful for consistency across answer buttons or feedback elements.",[12,1585,1586,1589,1590,1567,1593,1596],{},[57,1587,1588],{},"Lottie animations."," You can import ",[665,1591,1592],{},".lottie",[665,1594,1595],{},".json"," animation files from LottieFiles or After Effects (via Bodymovin), control them via triggers, and use the Animation Complete trigger to fire subsequent actions when a sequence finishes.",[12,1598,1599,1602,1603,1606,1607,1610],{},[57,1600,1601],{},"AI generation."," The fastest path to an Interactive Slides build is the ",[57,1604,1605],{},"Generate with AI"," button. You describe the simulation you want — a phishing email checker, a compliance decision tree, a software walkthrough — and the AI generates the full slideshow including elements, triggers, actions, and scoring. It is not always production-ready, but it gives you a working skeleton to refine rather than a blank canvas to start from. The ",[57,1608,1609],{},"Help Me Plan"," option is worth using before generating: it asks clarifying questions about scope, audience, and interaction style, which improves the output meaningfully.",[24,1612,1614],{"id":1613},"interactive-video-quiz-and-text-cuepoints-embedded-in-the-timeline","Interactive video: quiz and text cuepoints embedded in the timeline",[12,1616,1617],{},"Video blocks in LearnBuilder support cuepoints — moments in the video timeline where additional content appears.",[12,1619,1620],{},"There are two types:",[12,1622,1623,1626],{},[57,1624,1625],{},"Text cuepoints"," overlay a callout or annotation at a specific timestamp. Useful for adding context, highlighting something on screen, or flagging a key point without interrupting playback.",[12,1628,1629,1632],{},[57,1630,1631],{},"Quiz cuepoints"," pause the video at a specific timestamp and present a question. The video resumes after the learner answers. This is a straightforward implementation of a technique with solid research support — embedding retrieval practice inside video content improves retention compared to end-of-video quizzes, because the question fires while the content is still active in working memory.",[12,1634,1635],{},"Cuepoints work on uploaded video files. YouTube and Vimeo embeds are also supported in video blocks, though cuepoints require uploaded files.",[24,1637,1639],{"id":1638},"talking-head-video-lip-synced-presenter-video-from-a-portrait-and-script","Talking Head video: lip-synced presenter video from a portrait and script",[12,1641,1642],{},"Talking Head generates a lip-synced video from a character portrait and a text script. You pick a portrait image (upload, stock, AI-generated, or a course character), write the script, choose a voice, and optionally add a scene variation prompt to change the background or setting between videos — useful for avoiding visual repetition across multiple videos in a course.",[12,1644,1645],{},"The system generates VTT subtitles automatically from the script, which means every generated video is accessible out of the box rather than requiring a manual caption step.",[12,1647,1648],{},"From an instructional design perspective, Talking Head is most useful for instructor-style presentations, scenario introductions, and character voices in content that doesn't need full AI Dialogue interactivity. It is faster than recording video and sidesteps the production overhead of camera, lighting, and editing — which matters when you are building a course with ten or fifteen video segments.",[24,1650,1652],{"id":1651},"custom-embed-htmlcssjavascript-when-nothing-else-fits","Custom Embed: HTML\u002FCSS\u002FJavaScript when nothing else fits",[12,1654,1655,1656,543],{},"The Custom Embed block lets you write or generate arbitrary HTML, CSS, and JavaScript that runs inside a sandboxed iframe within the lesson. You can also upload files (images, data files) and reference them in code via ",[665,1657,1658],{},"EmbedFiles.get('filename')",[12,1660,1661],{},"AI generation is available here too: describe the activity in plain language and the AI writes the code. You can iterate with follow-up prompts to refine. For IDs who are not developers, this is a useful capability — the AI can produce working interactive content from a clear description.",[12,1663,1664],{},"Where Custom Embed earns its place is for things that do not fit the standard block types: custom calculators, branded interactive infographics, mini-games, mock interfaces for software training, or any interaction with unusual requirements. It is an escape hatch that experienced designers occasionally need.",[24,1666,1668],{"id":1667},"ai-assessment-grading-open-ended-responses-at-scale","AI Assessment: grading open-ended responses at scale",[12,1670,1671],{},"The Assessment Agent grades short-answer questions, essay responses, and AI Dialogue conversations. For quiz questions, you define a model answer and criteria; the agent evaluates each learner's response against them and provides written feedback.",[12,1673,1674],{},"For IDs, this changes the cost calculation on open-ended practice. The main reason knowledge checks in most courses are multiple-choice is not that multiple-choice is pedagogically ideal — it is that open-ended responses require a human to read and grade. The Assessment Agent makes it practical to include short-answer questions in a self-paced course and have every learner receive individualized written feedback. That is a meaningful capability for courses where the learning objective is something more than fact recall.",[24,1676,1678],{"id":1677},"accessibility-checker","Accessibility checker",[12,1680,1681],{},"The lesson editor includes a built-in accessibility checker that flags issues against WCAG standards before you publish. It catches missing captions on video blocks, missing alt text on images, hover-only interactions that are not keyboard accessible, and low-contrast elements, among others. Each issue links to the relevant WCAG rule.",[12,1683,1684],{},"For designers working to accessibility requirements — whether regulatory or organizational — having this built in is more useful than checking after the fact.",[24,1686,1688],{"id":1687},"scorm-xapi-import","SCORM \u002F xAPI import",[12,1690,1691],{},"If you have existing content in Storyline, Rise, iSpring, or any other tool that exports a valid SCORM or xAPI package, the SCORM\u002FxAPI block imports it directly. The package runs inside the lesson and completion and scores are recorded in the learner dashboard.",[12,1693,1694],{},"This matters practically because migration is never all-or-nothing. You may have a library of Storyline content you are not going to rebuild from scratch, alongside new content you are building in LearnBuilder. The import block means those can coexist in the same course.",[24,1696,1698],{"id":1697},"what-ai-actually-does-in-learnbuilder-for-experienced-designers","What AI actually does in LearnBuilder for experienced designers",[12,1700,1701],{},"Worth being explicit about this, because \"AI-powered authoring tool\" has become a category claim that covers everything from genuine capability to marketing language.",[12,1703,1704],{},"In LearnBuilder, AI does three things that are genuinely useful if you already know what you are doing:",[12,1706,1707,1710],{},[57,1708,1709],{},"It accelerates the build, not the design."," The Creator Agent can generate a course outline, write lesson content, create images, produce talking-head videos, and build Interactive Slides simulations from a description. None of that replaces instructional analysis or learning objective design — it starts from what you give it. If you feed it well-structured learning objectives and good scenario context, you get a usable first draft. If you feed it vague inputs, you get vague output.",[12,1712,1713,1716],{},[57,1714,1715],{},"It enables things that would otherwise require significant production overhead."," AI Dialogue scenarios and Talking Head videos require resources (writers, actors, video production) at scale in a traditional production process. The AI equivalent is not the same as a professionally produced scenario — but it is closer than nothing, available in minutes, and iterable.",[12,1718,1719,1722],{},[57,1720,1721],{},"It grades what you cannot grade at scale."," The Assessment Agent makes open-ended practice genuinely viable in a self-paced course. That changes what you can include without inflating review time.",[211,1724],{},[12,1726,1727,1728,1731],{},"If you are an instructional designer evaluating whether LearnBuilder fits your practice, the ",[171,1729,223],{"href":220,"rel":1730},[222]," is the best place to form a view. The AI Dialogue block and Interactive Slides builder in particular are the features that tend to answer the question — they are where the depth of the tool shows up most clearly.",{"title":226,"searchDepth":227,"depth":227,"links":1733},[1734,1735,1736,1737,1738,1739,1740,1741,1742],{"id":1478,"depth":227,"text":1479},{"id":1529,"depth":227,"text":1530},{"id":1613,"depth":227,"text":1614},{"id":1638,"depth":227,"text":1639},{"id":1651,"depth":227,"text":1652},{"id":1667,"depth":227,"text":1668},{"id":1677,"depth":227,"text":1678},{"id":1687,"depth":227,"text":1688},{"id":1697,"depth":227,"text":1698},"2026-04-14","Most of what gets written about AI authoring tools is aimed at people who have never built a course before. This post is for the people who have — and who want to know whether LearnBuilder has enough depth to be worth their time.","\u002Fblog\u002Flearnbuilder-for-instructional-designers.webp",{},"\u002Fblog\u002Flearnbuilder-for-instructional-designers",{"title":1458,"description":1744},"blog\u002Flearnbuilder-for-instructional-designers",[250,1454],"jwrkWETd4uZBPXGdnlalMM9vC-8NcZgd7om_ll8dZBg",{"id":1753,"title":1754,"author":7,"body":1755,"date":2301,"description":2302,"extension":239,"image":2303,"meta":2304,"navigation":242,"path":2305,"seo":2306,"stem":2307,"tags":2308,"__hash__":2310},"blog\u002Fblog\u002Ftop-10-elearning-authoring-tools-for-2026.md","Top 10 eLearning Authoring Tools for 2026",{"type":9,"value":1756,"toc":2283},[1757,1760,1763,1767,1770,1777,1781,1786,1789,1796,1799,1802,1807,1827,1833,1842,1844,1848,1851,1854,1857,1861,1878,1883,1888,1890,1894,1897,1900,1907,1911,1928,1933,1938,1940,1944,1947,1950,1953,1957,1974,1979,1984,1986,1990,1993,1996,1999,2003,2017,2022,2027,2029,2033,2036,2039,2042,2046,2060,2065,2070,2072,2076,2079,2082,2085,2089,2103,2108,2113,2115,2119,2122,2125,2128,2132,2146,2151,2156,2158,2162,2165,2168,2171,2175,2189,2194,2199,2201,2205,2208,2211,2215,2229,2234,2239,2241,2245,2252,2255,2258,2261,2264,2268,2271,2278],[12,1758,1759],{},"So this list has a point of view. I'll be upfront about it: I built LearnBuilder, and it sits at number one. But I've also used or thoroughly evaluated every tool on this list, and I've tried to be honest about what each one does well — and where it falls short.",[12,1761,1762],{},"If you're looking for a tool that helps you finish a check-the-box training, almost everything here will work. If you're looking for a tool that changes behavior on the job, the list gets shorter.",[24,1764,1766],{"id":1765},"what-are-elearning-authoring-tools","What Are eLearning Authoring Tools?",[12,1768,1769],{},"eLearning authoring tools are software platforms used to design, develop, and publish digital learning content — courses, assessments, simulations, and other materials — for delivery through a Learning Management System (LMS) or directly via a browser.",[12,1771,1772,1773,1776],{},"Most tools today support SCORM and xAPI standards for LMS integration, offer some degree of multimedia authoring, and have added AI features for content generation. Where they differ significantly is in ",[16,1774,1775],{},"how"," they support the instructional design process and how much they help you build content that actually drives learning outcomes — not just content that looks good on a completion report.",[24,1778,1780],{"id":1779},"the-top-10-elearning-authoring-tools-for-2026","The Top 10 eLearning Authoring Tools for 2026",[1782,1783,1785],"h3",{"id":1784},"_1-learnbuilder","1. LearnBuilder",[12,1787,1788],{},"LearnBuilder is built around a principle most authoring tools ignore: content that gets clicked through isn't the same as content that gets learned. Every design decision in the platform traces back to learning science — specifically the research on what actually changes behavior on the job.",[12,1790,1791,1792,1795],{},"Where other tools offer interactivity as a feature, LearnBuilder treats it as a question: ",[16,1793,1794],{},"does this interaction serve the learning objective, or just break up the monotony?"," Scenario-based learning, retrieval practice, spaced repetition, and dialogue simulations aren't add-ons here — they're the core of how courses are structured.",[12,1797,1798],{},"The AI in LearnBuilder is built for instructional design, not just content generation. It helps you scaffold scenarios, build realistic dialogue, generate retrieval practice questions, and structure content for long-term retention. The constraint isn't the AI — it's that building complex, realistic simulations requires you to think carefully about what real workplace decisions look like. LearnBuilder makes that thinking easier, but it doesn't skip it.",[12,1800,1801],{},"The built-in AI tutor gives learners an always-available resource trained on the course content itself — reducing dependency on instructors while keeping learners engaged after the formal module ends. SCORM, xAPI, and LTI integration means you can deploy into whatever LMS you're already using.",[12,1803,1804],{},[57,1805,1806],{},"Key Features:",[455,1808,1809,1812,1815,1818,1821,1824],{},[458,1810,1811],{},"Scenario-based learning and role play authoring",[458,1813,1814],{},"Retrieval practice and spaced repetition built into the course structure",[458,1816,1817],{},"Dialogue simulations for soft skills and decision-making training",[458,1819,1820],{},"AI tutor trained on your course content",[458,1822,1823],{},"SCORM, xAPI, and LTI support",[458,1825,1826],{},"Browser-based authoring — no software to install",[12,1828,1829,1832],{},[57,1830,1831],{},"Best For:"," Instructional designers, HR managers, and L&D teams who want courses that measurably change how people work — not just how they score on a completion report.",[12,1834,1835,1838,1839,543],{},[57,1836,1837],{},"Pricing:"," Free trial available at ",[171,1840,1276],{"href":220,"rel":1841},[222],[211,1843],{},[1782,1845,1847],{"id":1846},"_2-articulate-360-rise-storyline","2. Articulate 360 (Rise + Storyline)",[12,1849,1850],{},"Articulate 360 is the industry standard, and for good reason. It contains two authoring tools with very different personalities: Rise 360 for fast, clean, responsive courses, and Storyline 360 for complex, highly customized interactivity. If you have experienced instructional designers and complex branching requirements, Storyline is genuinely hard to beat.",[12,1852,1853],{},"The recent addition of AI features accelerates content drafting, and the Content Library 360 gives you access to millions of royalty-free assets. The collaboration and review workflow (Review 360) is mature and widely adopted.",[12,1855,1856],{},"The honest limitation: it's expensive, especially at team scale. And Rise — the tool most users actually spend their time in — produces content that tends to look and feel similar across organizations. If your learners see a lot of Rise courses, the format becomes invisible rather than engaging.",[12,1858,1859],{},[57,1860,1806],{},[455,1862,1863,1866,1869,1872,1875],{},[458,1864,1865],{},"Storyline 360 for custom branching, triggers, and advanced interactions",[458,1867,1868],{},"Rise 360 for rapid responsive course creation",[458,1870,1871],{},"AI Assistant for content drafting, image generation, and quiz creation",[458,1873,1874],{},"Content Library 360 with 9M+ royalty-free assets",[458,1876,1877],{},"Review 360 for structured stakeholder feedback",[12,1879,1880,1882],{},[57,1881,1831],{}," Organizations with experienced instructional designers who need maximum creative control, or teams already embedded in the Articulate ecosystem.",[12,1884,1885,1887],{},[57,1886,1837],{}," From $1,199\u002Fuser\u002Fyear (Personal) to $1,749\u002Fuser\u002Fyear (AI Teams plan).",[211,1889],{},[1782,1891,1893],{"id":1892},"_3-ispring-suite","3. iSpring Suite",[12,1895,1896],{},"iSpring's superpower is its PowerPoint integration. If your organization already lives in PowerPoint, iSpring Suite turns those decks into proper eLearning content — with quizzes, video narration, role-play simulations, and SCORM export — without forcing a new authoring workflow on your team.",[12,1898,1899],{},"The dialogue simulation builder is a standout feature for soft skills training. It's not as flexible as building branching scenarios from scratch, but for onboarding, compliance, and communication skills, it gets you far with relatively little friction.",[12,1901,1902,1903,1906],{},"The trade-off is that courses often ",[16,1904,1905],{},"look"," like they started in PowerPoint, even with polish applied. It's a rapid-authoring tool, and the ceiling on design quality and interactivity is lower than more purpose-built platforms.",[12,1908,1909],{},[57,1910,1806],{},[455,1912,1913,1916,1919,1922,1925],{},[458,1914,1915],{},"Deep PowerPoint integration for rapid content conversion",[458,1917,1918],{},"Dialogue simulation builder for role-play scenarios",[458,1920,1921],{},"Built-in video editor and screen recording",[458,1923,1924],{},"Full SCORM, xAPI, and cmi5 compliance",[458,1926,1927],{},"24\u002F7 technical support on all plans",[12,1929,1930,1932],{},[57,1931,1831],{}," Teams with strong PowerPoint skills who need to produce training quickly from existing slide-based content.",[12,1934,1935,1937],{},[57,1936,1837],{}," From $770\u002Fuser\u002Fyear (Suite) to $970\u002Fuser\u002Fyear (Suite Max, which adds AI features and a content library).",[211,1939],{},[1782,1941,1943],{"id":1942},"_4-coursebox","4. Coursebox",[12,1945,1946],{},"Coursebox is a good example of where AI-first authoring tools have arrived in 2026: genuinely fast content generation, a reasonable LMS included, and a low barrier to entry. Upload a document, a video, or a website URL and Coursebox will produce a structured course draft in minutes — including quizzes, flashcards, and an AI chatbot tutor.",[12,1948,1949],{},"The speed is real. For training providers, onboarding content, or organizations without dedicated L&D resources, Coursebox gets you to a working first draft faster than almost anything else on this list.",[12,1951,1952],{},"The limitation — and it's an important one — is that fast generation isn't the same as good instructional design. The AI produces content efficiently, but whether that content is structured for retention and meaningful practice depends heavily on the person editing it. Coursebox makes the mechanics easy; the design thinking is still on you.",[12,1954,1955],{},[57,1956,1806],{},[455,1958,1959,1962,1965,1968,1971],{},[458,1960,1961],{},"AI course creator from documents, videos, and website URLs",[458,1963,1964],{},"AI-generated quizzes and assessments",[458,1966,1967],{},"Built-in AI tutor chatbot trained on course content",[458,1969,1970],{},"Branded white-label LMS option",[458,1972,1973],{},"Supports 100+ languages",[12,1975,1976,1978],{},[57,1977,1831],{}," Training providers and organizations without dedicated L&D teams who need to produce content quickly and at low cost.",[12,1980,1981,1983],{},[57,1982,1837],{}," Free plan available. Creator plan from $29.99\u002Fmonth; Creator Plus at $99.99\u002Fmonth.",[211,1985],{},[1782,1987,1989],{"id":1988},"_5-easygenerator","5. Easygenerator",[12,1991,1992],{},"Easygenerator was built around a specific insight: subject matter experts, not instructional designers, often hold the knowledge organizations need to capture. The platform is intentionally simple — clean interface, pre-built templates, cloud-based — so that a sales manager or product expert can author a course without a week of onboarding.",[12,1994,1995],{},"The employee-generated content model has real merits in contexts where knowledge is dispersed and fast-moving. The review and feedback workflow makes it easy for L&D to maintain quality control without becoming a bottleneck.",[12,1997,1998],{},"The trade-off: simplicity means limits. Easygenerator won't take you far if your learning objectives require complex branching, high-fidelity simulations, or deep interactivity. It's optimized for getting knowledge out of people's heads quickly, not for engineering behavior change.",[12,2000,2001],{},[57,2002,1806],{},[455,2004,2005,2008,2011,2014],{},[458,2006,2007],{},"Designed for subject matter experts, not just L&D professionals",[458,2009,2010],{},"Built-in review and feedback tools for L&D quality control",[458,2012,2013],{},"Cloud-based collaborative authoring",[458,2015,2016],{},"Clean, template-driven interface",[12,2018,2019,2021],{},[57,2020,1831],{}," Organizations adopting an employee-generated content strategy, or L&D teams who need to scale content creation beyond their own department.",[12,2023,2024,2026],{},[57,2025,1837],{}," From $99\u002Fuser\u002Fmonth.",[211,2028],{},[1782,2030,2032],{"id":2031},"_6-lectora","6. Lectora",[12,2034,2035],{},"Lectora has been around long enough that its reputation is built on substance: it is the best-in-class tool for accessibility and regulatory compliance. If your organization operates in regulated industries, serves learners with disabilities, or needs to certify WCAG compliance, Lectora should be near the top of your list.",[12,2037,2038],{},"The built-in accessibility checker catches issues other tools leave for the developer to find manually. Multi-device publishing is solid, and the SCORM\u002FxAPI support is mature and reliable.",[12,2040,2041],{},"Where Lectora lags is in the authoring experience itself — it hasn't modernized as aggressively as competitors, and the interface can feel dated compared to Rise or iSpring. For teams where accessibility is a primary requirement rather than an afterthought, that trade-off is worth making.",[12,2043,2044],{},[57,2045,1806],{},[455,2047,2048,2051,2054,2057],{},[458,2049,2050],{},"Integrated accessibility checker for WCAG compliance",[458,2052,2053],{},"Multi-device publishing",[458,2055,2056],{},"SCORM and xAPI support",[458,2058,2059],{},"Mature compliance and regulatory training features",[12,2061,2062,2064],{},[57,2063,1831],{}," Organizations in regulated industries (healthcare, government, financial services) or any team with significant accessibility requirements.",[12,2066,2067,2069],{},[57,2068,1837],{}," From approximately $999\u002Fuser\u002Fyear.",[211,2071],{},[1782,2073,2075],{"id":2074},"_7-h5p","7. H5P",[12,2077,2078],{},"H5P occupies a unique position on this list: it's free, open-source, and browser-based, and supports a genuinely impressive range of interactive content types — branching scenarios, interactive video, drag-and-drop, flashcards, memory games, and more.",[12,2080,2081],{},"For educational institutions, non-profits, or organizations already running Moodle, WordPress, or Drupal, H5P integrates cleanly and adds meaningful interactivity without a licensing cost. The development community is active and the content type library keeps growing.",[12,2083,2084],{},"The limitation is that H5P is a content type system, not a full authoring environment. Building a coherent course requires more manual orchestration than polished tools. And the ceiling on visual quality and brand consistency is lower than purpose-built platforms.",[12,2086,2087],{},[57,2088,1806],{},[455,2090,2091,2094,2097,2100],{},[458,2092,2093],{},"Free and open-source",[458,2095,2096],{},"50+ interactive content types including branching scenarios",[458,2098,2099],{},"Integrates with Moodle, WordPress, Drupal, and Canvas",[458,2101,2102],{},"Active development community",[12,2104,2105,2107],{},[57,2106,1831],{}," Educational institutions, non-profits, and organizations with existing Moodle or WordPress infrastructure who need cost-effective interactive content.",[12,2109,2110,2112],{},[57,2111,1837],{}," Free (self-hosted). H5P.com hosted plans from $57\u002Fmonth.",[211,2114],{},[1782,2116,2118],{"id":2117},"_8-adobe-captivate","8. Adobe Captivate",[12,2120,2121],{},"Adobe Captivate is the specialist on this list. If you're building VR simulations, 360-degree immersive experiences, or software simulation training, Captivate has capabilities that most competitors simply don't offer.",[12,2123,2124],{},"The drag-and-drop editor is accessible to non-programmers, and the responsive design engine handles mobile well. For organizations where simulated environments or system walkthroughs are core to the training — manufacturing, healthcare, IT — Captivate earns its place.",[12,2126,2127],{},"For general corporate training, it's probably more tool than you need. The learning curve is steeper than Rise or Easygenerator, and for straightforward content you're paying for capabilities you won't use.",[12,2129,2130],{},[57,2131,1806],{},[455,2133,2134,2137,2140,2143],{},[458,2135,2136],{},"VR and 360-degree course creation",[458,2138,2139],{},"Software simulation and screen recording",[458,2141,2142],{},"Responsive design for mobile-first delivery",[458,2144,2145],{},"Drag-and-drop editor with no coding required",[12,2147,2148,2150],{},[57,2149,1831],{}," Organizations investing in immersive learning — VR simulations, software training, or technical procedure walkthroughs.",[12,2152,2153,2155],{},[57,2154,1837],{}," From $33.99\u002Fmonth (annual subscription).",[211,2157],{},[1782,2159,2161],{"id":2160},"_9-mindsmith","9. Mindsmith",[12,2163,2164],{},"Mindsmith is an AI-native authoring tool that has moved quickly to integrate AI into every step of the production workflow — from outline generation to content drafting to SCORM publishing. For small L&D teams or individual designers under deadline pressure, the efficiency gains are real.",[12,2166,2167],{},"The collaboration features are well-implemented, and the dynamic SCORM output handles LMS publishing without the friction some tools introduce. It's a modern tool with a modern interface.",[12,2169,2170],{},"Where it sits in the landscape: Mindsmith is primarily a productivity accelerator for experienced designers, not a framework for instructional quality. The AI helps you produce content faster; the quality of that content still depends on the design thinking you bring to it. If production speed is the primary constraint, it competes well.",[12,2172,2173],{},[57,2174,1806],{},[455,2176,2177,2180,2183,2186],{},[458,2178,2179],{},"AI integrated throughout the authoring workflow",[458,2181,2182],{},"Real-time team collaboration",[458,2184,2185],{},"Dynamic SCORM for LMS publishing",[458,2187,2188],{},"Multilingual course version management",[12,2190,2191,2193],{},[57,2192,1831],{}," Designers and small teams prioritizing production speed and modern collaboration workflows.",[12,2195,2196,2198],{},[57,2197,1837],{}," From $39\u002Fmonth.",[211,2200],{},[1782,2202,2204],{"id":2203},"_10-dominknow-one","10. dominKnow | ONE",[12,2206,2207],{},"dominKnow | ONE sits at the enterprise end of the spectrum. It combines responsive course authoring with software simulation, screen capture, and a centralized content management system. For large organizations managing training across many markets and languages, the multi-author collaboration and content reuse capabilities reduce overhead meaningfully.",[12,2209,2210],{},"The screen recording and software simulation tools are genuinely strong — comparable to Captivate for IT and systems training. The platform is more complex to onboard than simpler tools, but the depth justifies the investment for teams producing high volumes of diverse content.",[12,2212,2213],{},[57,2214,1806],{},[455,2216,2217,2220,2223,2226],{},[458,2218,2219],{},"Multi-author real-time collaboration",[458,2221,2222],{},"Screen recording and software simulation",[458,2224,2225],{},"Centralized content management for version control and reuse",[458,2227,2228],{},"Responsive authoring for all devices",[12,2230,2231,2233],{},[57,2232,1831],{}," Large enterprises managing high-volume, multi-market training programs — especially those requiring software simulation alongside other content types.",[12,2235,2236,2238],{},[57,2237,1837],{}," From $997\u002Fauthor\u002Fyear.",[211,2240],{},[24,2242,2244],{"id":2243},"how-to-choose-the-right-authoring-tool","How to Choose the Right Authoring Tool",[12,2246,2247,2248,2251],{},"Before looking at features or pricing, it's worth asking a more fundamental question: ",[16,2249,2250],{},"what does success actually look like for your training?"," The answer shapes everything else.",[12,2253,2254],{},"If your goal is compliance completion, almost any tool on this list will get you there. Choose based on budget and ease of use.",[12,2256,2257],{},"If your goal is behavior change — if you need salespeople to handle objections differently, or managers to give better feedback — you need a tool that supports practice-based learning, not just content consumption. That narrows the list considerably.",[12,2259,2260],{},"If you have dedicated instructional designers, tools like Articulate 360 or dominKnow offer deep creative control. If you don't, something with built-in instructional scaffolding (LearnBuilder) or a very low floor for subject matter experts (Easygenerator, Coursebox) will serve you better.",[12,2262,2263],{},"On pricing: the sticker price rarely reflects the full cost. Factor in the time your designers spend on production, the cost of content that doesn't achieve its learning objective, and — if you're evaluating legacy tools — the annual license renewal whether you use it actively or not.",[24,2265,2267],{"id":2266},"final-thoughts","Final Thoughts",[12,2269,2270],{},"The authoring tool market has never been more crowded. AI has compressed the distance between a blank page and a published course to minutes. That's genuinely useful — but it also means the differentiator is no longer speed of production. It's whether what you produce actually works.",[12,2272,2273,2274,2277],{},"If you want to explore what it looks like to build training around retention and behavior change — rather than just content delivery — ",[171,2275,572],{"href":220,"rel":2276},[222]," is worth a look. Free trial available, no credit card required.",[12,2279,2280],{},[16,2281,2282],{},"Have a specific use case you'd like to talk through? Reach out directly through the site.",{"title":226,"searchDepth":227,"depth":227,"links":2284},[2285,2286,2299,2300],{"id":1765,"depth":227,"text":1766},{"id":1779,"depth":227,"text":1780,"children":2287},[2288,2290,2291,2292,2293,2294,2295,2296,2297,2298],{"id":1784,"depth":2289,"text":1785},3,{"id":1846,"depth":2289,"text":1847},{"id":1892,"depth":2289,"text":1893},{"id":1942,"depth":2289,"text":1943},{"id":1988,"depth":2289,"text":1989},{"id":2031,"depth":2289,"text":2032},{"id":2074,"depth":2289,"text":2075},{"id":2117,"depth":2289,"text":2118},{"id":2160,"depth":2289,"text":2161},{"id":2203,"depth":2289,"text":2204},{"id":2243,"depth":227,"text":2244},{"id":2266,"depth":227,"text":2267},"2026-04-11","A practical comparison of the best e-learning authoring tools in 2026 — from AI-powered course creators to traditional SCORM editors. What's changed, what matters, and which tool fits your workflow.","\u002Fblog\u002Ftop-10-elearning-authoring-tools-2026.webp",{},"\u002Fblog\u002Ftop-10-elearning-authoring-tools-for-2026",{"title":1754,"description":2302},"blog\u002Ftop-10-elearning-authoring-tools-for-2026",[1454,2309,813],"comparison","fIYRa-WQxSyojJC1ufHejkMdOnuoCvCbDNimZtELWG0",{"id":2312,"title":2313,"author":7,"body":2314,"date":2519,"description":2520,"extension":239,"image":2521,"meta":2522,"navigation":242,"path":2523,"seo":2524,"stem":2525,"tags":2526,"__hash__":2530},"blog\u002Fblog\u002Fmicrolearning-authoring-with-AI.md","Microlearning authoring with AI: bite-sized lessons that stick",{"type":9,"value":2315,"toc":2511},[2316,2319,2322,2325,2328,2330,2334,2337,2340,2343,2346,2348,2352,2359,2362,2376,2379,2381,2385,2388,2391,2397,2403,2409,2415,2417,2421,2424,2427,2445,2448,2450,2454,2457,2463,2469,2475,2481,2483,2487,2494,2497,2500,2502],[12,2317,2318],{},"There's a pattern that comes up constantly in workplace training: someone builds a comprehensive, well-structured course. It covers everything. It takes 90 minutes to complete. Nobody finishes it.",[12,2320,2321],{},"The training team wonders what went wrong. The answer is usually not the content — it's the format.",[12,2323,2324],{},"People don't learn well in 90-minute blocks, especially not at work. Attention drifts. Notifications compete. The course gets bookmarked, then forgotten. Meanwhile, the gap between what people know and what they need to know stays open.",[12,2326,2327],{},"Microlearning is the practical response to this reality.",[211,2329],{},[24,2331,2333],{"id":2332},"what-microlearning-actually-is","What microlearning actually is",[12,2335,2336],{},"Microlearning isn't just \"short videos\" or \"bite-sized content\" — though those can be part of it. It's a design principle: structure learning around single, specific skills or concepts, make each unit completeable in 3–10 minutes, and focus on immediate application.",[12,2338,2339],{},"A good microlearning unit answers one question. It doesn't try to be comprehensive. It doesn't include background context that isn't directly relevant. It gets in, delivers one actionable thing, and gets out.",[12,2341,2342],{},"This sounds simple. In practice it's harder than building long-form courses, because it requires ruthless prioritisation. Every sentence has to earn its place. There's no room to hedge or over-explain.",[12,2344,2345],{},"That difficulty is exactly why AI changes the equation.",[211,2347],{},[24,2349,2351],{"id":2350},"the-authoring-problem-with-microlearning-at-scale","The authoring problem with microlearning at scale",[12,2353,2354,2355,2358],{},"A single microlearning unit is easy enough to build. The challenge is that effective microlearning programmes are made of ",[16,2356,2357],{},"many"," units — often 20, 40, or more, covering different aspects of a topic in a structured sequence.",[12,2360,2361],{},"Building that manually is a significant undertaking, even with modern authoring tools. You need to:",[455,2363,2364,2367,2370,2373],{},[458,2365,2366],{},"Break a topic into discrete, non-overlapping skills",[458,2368,2369],{},"Write focused content for each one (no padding, no repetition)",[458,2371,2372],{},"Create scenarios and questions that test application, not recall",[458,2374,2375],{},"Keep a consistent voice and structure across all units",[12,2377,2378],{},"For a single course developer, this is weeks of work. For a non-specialist — an HR manager, a compliance officer, an operations lead — it's often simply not feasible.",[211,2380],{},[24,2382,2384],{"id":2383},"where-ai-actually-helps","Where AI actually helps",[12,2386,2387],{},"AI doesn't replace judgment in course design. It does remove the parts that take the most time without requiring much judgment: generating the first draft, turning a policy document into lesson content, writing scenario-based questions, and filling in explanatory text.",[12,2389,2390],{},"In practice, this means:",[12,2392,2393,2396],{},[57,2394,2395],{},"Faster breakdowns."," Given a topic or a source document, AI can quickly suggest how to break it into microlearning units — which skills are discrete enough to stand alone, what sequence makes sense, where prerequisites exist.",[12,2398,2399,2402],{},[57,2400,2401],{},"Draft content in minutes."," Instead of staring at a blank page for each 5-minute lesson, you start with a draft that covers the key point. Your job shifts from writing to editing — which is significantly faster.",[12,2404,2405,2408],{},[57,2406,2407],{},"Better scenarios."," The hardest part of microlearning isn't the explanatory content — it's writing realistic scenarios that put the learner in a situation where they have to apply what they've just learned. AI is genuinely good at generating first drafts of these.",[12,2410,2411,2414],{},[57,2412,2413],{},"Consistency across units."," When you're building 30 units, keeping the tone and structure consistent is surprisingly hard. Working with AI helps maintain that consistency without having to manually check each unit against the others.",[211,2416],{},[24,2418,2420],{"id":2419},"what-this-looks-like-in-practice","What this looks like in practice",[12,2422,2423],{},"Imagine you're an HR manager who needs to train 50 people on updated data handling procedures. Previously your options were: write a long document, run a meeting, or commission a course that would take months to develop.",[12,2425,2426],{},"With an AI-assisted authoring tool, a realistic workflow looks like this:",[2428,2429,2430,2433,2436,2439,2442],"ol",{},[458,2431,2432],{},"Paste in the relevant policy sections",[458,2434,2435],{},"Get a suggested breakdown of 8–12 microlearning units covering distinct aspects of the policy",[458,2437,2438],{},"Review and edit the draft content for each unit (15–20 minutes per unit)",[458,2440,2441],{},"Add a short scenario question to each unit",[458,2443,2444],{},"Publish",[12,2446,2447],{},"The whole process — for a substantive topic — can happen in an afternoon. The result isn't a document people skim. It's a structured learning experience they can complete in short sessions, on whatever device they have available.",[211,2449],{},[24,2451,2453],{"id":2452},"the-design-principles-that-still-matter","The design principles that still matter",[12,2455,2456],{},"AI handles speed and volume. The judgment calls are still yours:",[12,2458,2459,2462],{},[57,2460,2461],{},"One skill per unit."," Resist the temptation to combine related concepts. If a learner needs to do two things differently after this unit, split it into two units.",[12,2464,2465,2468],{},[57,2466,2467],{},"Application, not information."," Information transfers in documents. Learning happens through application. Every microlearning unit should end with a scenario or decision point — not just a summary.",[12,2470,2471,2474],{},[57,2472,2473],{},"No dead content."," In a 5-minute unit, every sentence matters. If a sentence doesn't directly support the learning objective, cut it. AI-generated drafts often need editing with this in mind.",[12,2476,2477,2480],{},[57,2478,2479],{},"Sequence deliberately."," Microlearning works best when units build on each other. Foundational concepts first, application later, edge cases at the end.",[211,2482],{},[24,2484,2486],{"id":2485},"who-this-is-for","Who this is for",[12,2488,2489,2490,2493],{},"Microlearning with AI authoring isn't primarily for large L&D teams who already have mature processes and professional tools. They'll adopt it too, but it's most transformative for people who currently ",[16,2491,2492],{},"don't"," create structured learning content — because the time and skill barriers previously made it impractical.",[12,2495,2496],{},"HR managers, compliance leads, operations teams, customer success managers, technical trainers without a design background: these are the people who have knowledge that needs to be learned, but who've never had a realistic path to turning it into genuine training.",[12,2498,2499],{},"That path now exists.",[211,2501],{},[12,2503,2504],{},[16,2505,2506,2507],{},"LearnBuilder is built for exactly this — fast, AI-assisted microlearning authoring that produces interactive, trackable courses without requiring an instructional design background. ",[171,2508,2510],{"href":220,"rel":2509},[222],"Try it free →",{"title":226,"searchDepth":227,"depth":227,"links":2512},[2513,2514,2515,2516,2517,2518],{"id":2332,"depth":227,"text":2333},{"id":2350,"depth":227,"text":2351},{"id":2383,"depth":227,"text":2384},{"id":2419,"depth":227,"text":2420},{"id":2452,"depth":227,"text":2453},{"id":2485,"depth":227,"text":2486},"2026-04-08","Long courses get skipped. Microlearning gets done. Here's how AI makes it practical to create short, focused lessons that actually change behaviour — without a team of instructional designers.","\u002Fuse-cases\u002Fsme.webp",{},"\u002Fblog\u002Fmicrolearning-authoring-with-ai",{"title":2313,"description":2520},"blog\u002Fmicrolearning-authoring-with-AI",[2527,2528,2529],"microlearning","ai","practical","CByu9PL-R8NjkpMacODqF6lviqVFEfXAlRqXVbaWF9I",{"id":2532,"title":2533,"author":7,"body":2534,"date":2669,"description":2670,"extension":239,"image":2671,"meta":2672,"navigation":242,"path":2673,"seo":2674,"stem":2675,"tags":2676,"__hash__":2678},"blog\u002Fblog\u002Flms-confusion-draft.md","Your LMS doesn't come with courses — and that's the gap nobody warns you about",{"type":9,"value":2535,"toc":2662},[2536,2539,2542,2545,2547,2551,2554,2561,2564,2566,2570,2573,2579,2585,2588,2590,2594,2597,2600,2602,2606,2609,2612,2619,2621,2625,2628,2648,2651,2653],[12,2537,2538],{},"A question came up on Reddit this week that I see variations of constantly: someone had just set up TalentLMS and was confused to find it empty. Where were the courses? Wasn't an LMS supposed to come with training?",[12,2540,2541],{},"It's a completely reasonable thing to expect. The name — Learning Management System — implies that learning is included. It isn't. An LMS is a delivery platform, not a content library. It manages courses; it doesn't create them.",[12,2543,2544],{},"This distinction matters because it changes what problem you actually need to solve.",[211,2546],{},[24,2548,2550],{"id":2549},"what-an-lms-does-and-doesnt-do","What an LMS does and doesn't do",[12,2552,2553],{},"An LMS gives you infrastructure: a place to host courses, track who has completed them, issue certificates, and manage learner enrolments. If you have courses, it's genuinely useful. If you don't, it's an empty building.",[12,2555,2556,2557,2560],{},"This trips up a lot of organisations — especially smaller ones — who buy or sign up for an LMS expecting it to solve their training problem. The LMS solves the ",[16,2558,2559],{},"delivery"," problem. The content problem is separate.",[12,2562,2563],{},"That content problem is harder. Creating courses from scratch requires either time, skills, or money — usually all three. Which is why so many organisations end up with an LMS full of PDF uploads and recorded webinars, and call it training.",[211,2565],{},[24,2567,2569],{"id":2568},"the-two-problems-people-confuse","The two problems people confuse",[12,2571,2572],{},"When someone says \"we need a training solution,\" they usually mean one of two things:",[12,2574,2575,2578],{},[57,2576,2577],{},"Problem 1: We need somewhere to host and track training.","\nThis is the LMS problem. Learners need to access courses, complete them, and have that tracked. An LMS solves this.",[12,2580,2581,2584],{},[57,2582,2583],{},"Problem 2: We need to create training content.","\nThis is the authoring problem. Someone needs to turn knowledge — a policy, a process, a set of skills — into a structured learning experience. An LMS doesn't touch this.",[12,2586,2587],{},"Most small and mid-sized organisations have Problem 2 but spend their energy solving Problem 1. They get the LMS in place and then stall when they realise the content still needs to be built.",[211,2589],{},[24,2591,2593],{"id":2592},"why-content-creation-has-felt-hard","Why content creation has felt hard",[12,2595,2596],{},"Until recently, building courses required either instructional design expertise or a steep learning curve with professional authoring tools. Tools like Articulate Storyline are genuinely powerful, but they were designed for dedicated L&D professionals. For an HR manager or operations lead who needs to build one compliance course, the investment in learning the tool rarely made sense.",[12,2598,2599],{},"The result was a bifurcated market: large organisations with L&D teams who could use professional tools, and everyone else who made do with documents and slide decks.",[211,2601],{},[24,2603,2605],{"id":2604},"whats-changed","What's changed",[12,2607,2608],{},"AI has materially changed the content creation side of this equation. Not in a \"replace the instructional designer\" way — but in a \"make it feasible for non-designers to build real courses\" way.",[12,2610,2611],{},"The part of course creation that used to require the most time and skill — structuring information, writing lesson content, generating questions and scenarios — can now be done much faster with the right tooling. An HR manager with a policy document can now go from source material to a structured, interactive course in hours rather than days.",[12,2613,2614,2615,2618],{},"This doesn't make the LMS irrelevant. If you need to deploy training to many people, track completions, and generate compliance reports, you still need one. But for organisations that primarily need to ",[16,2616,2617],{},"create"," training — and especially for smaller teams who don't have a dedicated L&D function — the authoring problem is now much more tractable.",[211,2620],{},[24,2622,2624],{"id":2623},"a-practical-starting-point","A practical starting point",[12,2626,2627],{},"If you're in the position of having an LMS but no courses, or needing to build training without an L&D background, the order of operations matters:",[2428,2629,2630,2636,2642],{},[458,2631,2632,2635],{},[57,2633,2634],{},"Start with what learners need to be able to do"," — not what they need to know. Behaviour change, not information transfer, is the goal.",[458,2637,2638,2641],{},[57,2639,2640],{},"Build the content first"," — identify your source material, structure it into short lessons, add questions and scenarios. This is the authoring step.",[458,2643,2644,2647],{},[57,2645,2646],{},"Then deploy"," — once you have a course, the LMS question is straightforward.",[12,2649,2650],{},"Getting these in the wrong order — buying the LMS first, hoping the content problem will solve itself — is how you end up with an expensive empty platform.",[211,2652],{},[12,2654,2655],{},[16,2656,2657,2658],{},"LearnBuilder is an authoring tool designed for exactly this gap: teams who need to create professional training courses without a dedicated L&D function. Upload your source document, and LearnBuilder structures it into lessons, generates quizzes and scenarios, and adds an AI tutor grounded in your content. ",[171,2659,2661],{"href":220,"rel":2660},[222],"See how it works →",{"title":226,"searchDepth":227,"depth":227,"links":2663},[2664,2665,2666,2667,2668],{"id":2549,"depth":227,"text":2550},{"id":2568,"depth":227,"text":2569},{"id":2592,"depth":227,"text":2593},{"id":2604,"depth":227,"text":2605},{"id":2623,"depth":227,"text":2624},"2026-04-07","Most organisations buy an LMS expecting a training solution. What they get is infrastructure. The courses still need to come from somewhere — and that's where the real work begins.","\u002Fblog\u002Flms-confusion-draft.webp",{},"\u002Fblog\u002Flms-confusion-draft",{"title":2533,"description":2670},"blog\u002Flms-confusion-draft",[2529,2677],"lms","kaGFEonySiPZBj0ZvxIJUNt7_Ms-lnjmH4gF68o8LZs",{"id":2680,"title":2681,"author":7,"body":2682,"date":2862,"description":2863,"extension":239,"image":2864,"meta":2865,"navigation":242,"path":2866,"seo":2867,"stem":2868,"tags":2869,"__hash__":2870},"blog\u002Fblog\u002Fhow-to-turn-a-document-into-staff-training.md","How to turn a document into staff training — without an L&D team",{"type":9,"value":2683,"toc":2854},[2684,2687,2691,2694,2697,2700,2709,2712,2714,2718,2721,2727,2733,2739,2745,2747,2751,2754,2763,2766,2771,2774,2779,2782,2787,2790,2792,2796,2799,2802,2805,2808,2811,2813,2817,2820,2823,2826,2829,2831,2835,2841,2844,2846],[12,2685,2686],{},"There is a situation that comes up constantly in small and mid-sized organisations, and almost never gets discussed in learning and development circles — because the people dealing with it are not in L&D at all. A policy has changed. A new process has been rolled out. A regulator requires staff to demonstrate they understand a procedure. And somewhere in a shared drive sits a 20, 40, or 80-page document that contains everything staff need to know. The person responsible for turning that document into training — usually an HR manager, an operations lead, or occasionally a volunteer coordinator — has no instructional design background, no authoring tool licence, and no budget for a contractor. What usually happens: the document gets emailed out with a note asking staff to read it. Nobody reads it. The checkbox gets ticked anyway. This post is about a better approach — one that does not require an L&D background, a large budget, or weeks of production time.",[24,2688,2690],{"id":2689},"why-just-read-the-document-fails","Why \"just read the document\" fails",[12,2692,2693],{},"The instinct to send a document and ask people to read it is understandable. The information is all there. Surely motivated adults can absorb it themselves.",[12,2695,2696],{},"The problem is not motivation. It is how memory works.",[12,2698,2699],{},"Reading a document once — even carefully — produces weak, short-lived retention. The material feels familiar immediately after reading, which creates a misleading sense of having learned it. A week later, most of the detail is gone. This is not a character flaw in your staff. It is how human memory behaves when information is encountered passively, once, without any retrieval demand.",[12,2701,2702,2703,2708],{},"The research on this is consistent and goes back decades. ",[171,2704,2707],{"href":2705,"rel":2706},"https:\u002F\u002Fpsycnet.apa.org\u002Frecord\u002F2006-05194-012",[222],"Roediger and Karpicke (2006)"," showed that a group who read material once and then tried to recall it — three times, without looking — retained significantly more after one week than a group who read the same material four times. The act of retrieval, not re-reading, is what consolidates memory.",[12,2710,2711],{},"For organisations, this has a practical implication: a 30-minute training course that makes staff actively engage with the content will outperform a 30-page document every time — not because the document is wrong, but because passive reading is a poor vehicle for durable learning.",[211,2713],{},[24,2715,2717],{"id":2716},"what-good-training-from-a-document-actually-looks-like","What good training from a document actually looks like",[12,2719,2720],{},"You do not need to be an instructional designer to understand the principles behind effective training. They are not complicated.",[12,2722,2723,2726],{},[57,2724,2725],{},"Break it into digestible units."," A 40-page policy document is not a course. It is source material. Good training extracts the parts that staff actually need to act on and organises them into short, focused lessons — ideally with one main idea per lesson.",[12,2728,2729,2732],{},[57,2730,2731],{},"Ask questions during, not just after."," The standard e-learning format — content first, quiz at the end — is better than nothing, but not by much. Retrieval practice is most effective when it is interspersed throughout the content, not saved for a final assessment. A question mid-lesson, before the answer has been re-stated, forces active recall and strengthens memory in a way that end-of-course quizzes do not.",[12,2734,2735,2738],{},[57,2736,2737],{},"Connect content to real situations."," Staff are more likely to retain information when it is attached to a recognisable scenario: a situation they might actually face, a decision they might have to make. A policy on data handling is abstract. A scenario in which a staff member receives a request to share a client record — and has to decide what to do — is concrete. The procedure becomes a decision, not a fact to memorise.",[12,2740,2741,2744],{},[57,2742,2743],{},"Allow people to ask questions."," One limitation of any fixed training content is that it cannot anticipate every question. Staff in different roles will encounter the material differently. Providing a way for learners to ask clarifying questions — and get answers grounded in the actual content, not generic advice — closes the gap between training and application.",[211,2746],{},[24,2748,2750],{"id":2749},"a-practical-process-for-non-designers","A practical process for non-designers",[12,2752,2753],{},"If you are responsible for turning a document into training and have no instructional design background, here is a process that works.",[12,2755,2756],{},[57,2757,2758,2759,2762],{},"Step 1: Identify what staff need to ",[16,2760,2761],{},"do",", not just know.",[12,2764,2765],{},"Read through the document and ask: what decisions will staff have to make as a result of this? What procedures will they need to follow? What mistakes are most likely, and what are the consequences? The answers to these questions are your learning objectives — the outcomes the training needs to produce. Everything else in the document is context.",[12,2767,2768],{},[57,2769,2770],{},"Step 2: Organise the content into short lessons around each objective.",[12,2772,2773],{},"A lesson does not need to be long. Five to ten minutes of content per lesson, focused on a single objective, is more effective than a single 45-minute module that covers everything. Group related objectives into modules if you have many, but resist the temptation to build one big comprehensive course. Shorter, focused lessons are easier to update, easier to complete, and easier to remember.",[12,2775,2776],{},[57,2777,2778],{},"Step 3: Add a question or scenario to each lesson.",[12,2780,2781],{},"Every lesson should include at least one moment where learners have to actively engage with the content — not just read it. This can be simple: a multiple-choice question that asks them to apply a rule to a situation, a short-answer prompt asking them to explain a process in their own words, or a brief scenario with a decision point. The format matters less than the cognitive demand: learners should have to think, not just recognise.",[12,2783,2784],{},[57,2785,2786],{},"Step 4: Consider what questions staff are likely to ask.",[12,2788,2789],{},"Think through the questions a new staff member might have after completing the training. Where is the ambiguity? Where might someone need to know how a rule applies to their specific situation? Anticipating these questions — and making sure the training addresses them, or that there is a way for learners to get answers — significantly increases the practical value of what you build.",[211,2791],{},[24,2793,2795],{"id":2794},"what-ai-can-and-cannot-do-here","What AI can and cannot do here",[12,2797,2798],{},"AI tools have made it meaningfully faster to turn a document into structured training content. What used to take days of manual work — reading a source document, extracting key concepts, drafting lesson content and questions — can now be done in a fraction of the time.",[12,2800,2801],{},"But it is worth being clear about what AI does well here, and where human judgment still matters.",[12,2803,2804],{},"AI is good at extracting and restructuring information from a source document, drafting lesson content that reflects what the document actually says, generating quiz questions and scenarios from that content, and answering learner questions about the material — if it is given the document as a knowledge source rather than relying on general training.",[12,2806,2807],{},"AI is not reliable when it operates without grounding. A general-purpose AI assistant asked to create a compliance course on your data protection policy will produce something plausible-sounding but not necessarily accurate to your specific policy. The difference between useful AI-generated training and unreliable AI-generated training is almost always whether the AI was given your actual document or asked to work from its own knowledge.",[12,2809,2810],{},"This is worth naming because many organisations are experimenting with AI for training right now, and the most common failure mode is using AI that is not grounded in the source material. The result is training that sounds authoritative but does not actually reflect what the document says.",[211,2812],{},[24,2814,2816],{"id":2815},"a-note-on-what-training-can-realistically-achieve","A note on what training can realistically achieve",[12,2818,2819],{},"It is worth being honest about the limits of any training course, however well designed.",[12,2821,2822],{},"Training works best when it is addressing a genuine knowledge or skill gap — when staff do not know something, or cannot yet do something, and learning is what closes that gap. It is less effective as a substitute for clear processes, well-designed systems, or adequate resources. If staff keep making errors in a particular area, the first question is not always \"do they need more training?\" It is worth asking whether the process itself is unclear, or whether the environment makes errors easy.",[12,2824,2825],{},"Training is also not a one-time event. A course completed once, however good, will produce some decay over time. The most durable learning comes from spaced repetition — returning to material at intervals — and from application in real situations. A well-designed course gets staff to a good starting point. What happens after the course matters too.",[12,2827,2828],{},"None of this is a reason not to invest in good training. It is a reason to be clear about what you are trying to achieve, and honest about whether training is the right lever.",[211,2830],{},[24,2832,2834],{"id":2833},"where-to-start","Where to start",[12,2836,2837,2838,2840],{},"If you have a document that needs to become training and are not sure where to start, the most useful first step is not to open an authoring tool. It is to read the document with one question in mind: what do staff need to be able to ",[16,2839,2761],{}," after this training that they cannot do now?",[12,2842,2843],{},"Answer that question clearly, and the rest of the process — however you build it — has a much better chance of producing something that works.",[211,2845],{},[12,2847,2848],{},[16,2849,2850,2851],{},"LearnBuilder is an e-learning authoring tool that can generate structured lesson content from uploaded documents, including quizzes, scenarios, and an AI tutor grounded in your source material. It is designed for instructional designers, but works equally well for HR and operations teams building training without a dedicated L&D function. ",[171,2852,2661],{"href":220,"rel":2853},[222],{"title":226,"searchDepth":227,"depth":227,"links":2855},[2856,2857,2858,2859,2860,2861],{"id":2689,"depth":227,"text":2690},{"id":2716,"depth":227,"text":2717},{"id":2749,"depth":227,"text":2750},{"id":2794,"depth":227,"text":2795},{"id":2815,"depth":227,"text":2816},{"id":2833,"depth":227,"text":2834},"2026-04-03","Turn a policy document or handbook into real staff training — no L&D background, authoring-tool licence, or contractor budget needed. A practical walkthrough.","\u002Fblog\u002Fdocument-into-staff-training.webp",{},"\u002Fblog\u002Fhow-to-turn-a-document-into-staff-training",{"title":2681,"description":2863},"blog\u002Fhow-to-turn-a-document-into-staff-training",[2529,250],"Ku0P0BOSflS3Ic8m6_EYEeRrz0yCCNHr_M2rNlKV6Ao",{"id":2872,"title":2873,"author":7,"body":2874,"date":3169,"description":3170,"extension":239,"image":3171,"meta":3172,"navigation":242,"path":3173,"seo":3174,"stem":3175,"tags":3176,"__hash__":3177},"blog\u002Fblog\u002Fthe-case-against-active-learning.md","The case against active learning",{"type":9,"value":2875,"toc":3161},[2876,2893,2895,2898,2901,2903,2907,2919,2922,2938,2945,2947,2951,2954,2961,2989,2992,2995,2997,3001,3004,3007,3010,3013,3016,3018,3022,3025,3031,3050,3062,3064,3068,3071,3074,3080,3086,3092,3098,3100,3102,3105,3108,3111,3113],[12,2877,2878],{},[16,2879,2880,2881,2886,2887,2892],{},"This is the third post in a series exploring the evidence behind learning design decisions. The ",[171,2882,2885],{"href":2883,"rel":2884},"https:\u002F\u002Flearnbuilder.org\u002Fblog\u002Fwhy-most-e-learning-fails",[222],"first post"," examined why the tell-and-test format underperforms and what retrieval practice research tells us about retention. The ",[171,2888,2891],{"href":2889,"rel":2890},"https:\u002F\u002Flearnbuilder.org\u002Fblog\u002Flearning-through-decision",[222],"second post"," looked at scenario-based design and productive failure. This post takes the opposite direction: what does the research say when it argues that active learning might not be worth the effort?",[211,2894],{},[12,2896,2897],{},"If you have read the first two posts in this series, you could be forgiven for thinking the debate is settled. Active learning beats passive delivery. Retrieval practice outperforms re-reading. Scenarios outperform slides. Build interactive e-learning, and good outcomes will follow.",[12,2899,2900],{},"The evidence does point that way. But there is a serious counter-tradition in learning research — one that has been running for over forty years — that complicates the picture in ways that are genuinely useful for anyone designing e-learning. Engaging with the counterarguments honestly makes for better design decisions than simply citing the meta-analyses that support what we already want to believe.",[211,2902],{},[24,2904,2906],{"id":2905},"the-great-media-debate","The great media debate",[12,2908,2909,2910,2918],{},"In 1983, educational psychologist Richard Clark published ",[171,2911,2914,2915],{"href":2912,"rel":2913},"https:\u002F\u002Fwww.uky.edu\u002F~gmswan3\u002F609\u002FClark_1983.pdf",[222],"a paper in the ",[16,2916,2917],{},"Review of Educational Research"," that has shaped the field ever since. His argument was blunt: media are mere delivery vehicles for instruction and do not influence learning. The analogy he used — that a delivery truck does not affect the nutritional value of the groceries it carries — became one of the most quoted lines in educational technology.",[12,2920,2921],{},"Clark's claim was based on decades of media comparison studies. Researchers had compared film to lectures, television to textbooks, computer-based training to classroom instruction. The consistent finding was what became known as the \"no significant difference\" phenomenon: when the instructional method was held constant, switching the delivery medium produced no meaningful change in outcomes.",[12,2923,2924,2925,2930,2931,2933,2934,2937],{},"Robert Kozma ",[171,2926,2929],{"href":2927,"rel":2928},"http:\u002F\u002Fwww.ala.org\u002Faasl\u002Fsites\u002Fala.org.aasl\u002Ffiles\u002Fcontent\u002Faaslpubsandjournals\u002Fslr\u002Fedchoice\u002FSLMQ_InfluenceofMediaonLearning_InfoPower.pdf",[222],"responded in 1994",", arguing that the question should not be whether media ",[16,2932,2761],{}," influence learning, but whether they ",[16,2935,2936],{},"will"," when properly designed. He proposed that certain media afford certain cognitive processes — that a simulation can support mental model-building in ways a textbook cannot, not because simulations are inherently better, but because they enable different kinds of interaction with the material.",[12,2939,2940,2941,2944],{},"The debate never fully resolved. But Clark's core position has proven remarkably durable. Even his critics have largely shifted from arguing that media directly cause learning to arguing that media ",[16,2942,2943],{},"enable"," certain methods that cause learning — which is a more nuanced version of Clark's original point. The truck does not change the nutrition. But some trucks have refrigeration, which lets you carry cargo that would spoil otherwise.",[211,2946],{},[24,2948,2950],{"id":2949},"what-no-significant-difference-actually-means","What \"no significant difference\" actually means",[12,2952,2953],{},"The no significant difference finding is sometimes treated as proof that all formats are equally effective, and therefore we should choose whatever is cheapest. This is a misreading — and an important one to correct.",[12,2955,2956,2957,2960],{},"What the studies consistently show is that when you take the ",[16,2958,2959],{},"same instructional method"," and deliver it through different media, outcomes do not change. A recorded lecture produces similar results to a live lecture. A PDF of slides produces similar results to clicking through those slides online. This is not surprising. The method is doing the work. The medium is carrying it.",[12,2962,2963,2964,2967,2968,2971,2972,2977,2978,2982,2983,2988],{},"But here is the point that matters for e-learning design: the no significant difference finding does not say that all ",[16,2965,2966],{},"methods"," are equal. It says that all ",[16,2969,2970],{},"media"," are equal when the method is held constant. The ",[171,2973,2976],{"href":2974,"rel":2975},"https:\u002F\u002Fwww.pnas.org\u002Fdoi\u002Fabs\u002F10.1073\u002Fpnas.1319030111",[222],"Freeman meta-analysis",", the ",[171,2979,2981],{"href":2705,"rel":2980},[222],"testing effect literature",", and the ",[171,2984,2987],{"href":2985,"rel":2986},"https:\u002F\u002Feducation.asu.edu\u002Fsites\u002Fg\u002Ffiles\u002Flitvpz656\u002Ffiles\u002Flcl\u002Fchiwylie2014icap_2.pdf",[222],"ICAP framework"," all demonstrate substantial differences between methods. Active retrieval outperforms passive re-reading. Constructive engagement outperforms clicking. Scenario-based practice outperforms information presentation.",[12,2990,2991],{},"The practical implication is precise and worth stating clearly. If you take a slide deck, record a narrator reading it, and publish it as an e-learning module, you have changed the medium without changing the method. Clark's research predicts — correctly — that this will produce no improvement. If you instead redesign the content to include embedded retrieval questions, decision points, and spaced practice, you have changed the method. The evidence predicts — also correctly — that this will produce measurable gains.",[12,2993,2994],{},"The medium is not the message. The method is.",[211,2996],{},[24,2998,3000],{"id":2999},"the-case-for-passive-learning-honestly-stated","The case for passive learning — honestly stated",[12,3002,3003],{},"There are legitimate arguments for passive delivery that deserve more than dismissal.",[12,3005,3006],{},"First, passive does not necessarily mean cognitively idle. A learner reading a well-structured text while actively thinking about how it applies to their work is engaging in deep processing — even though they are not clicking, dragging, or selecting anything. Research on generative processing suggests that what matters is internal cognitive activity, not visible behavioural activity. A learner can be overtly passive while experiencing high levels of cognitive engagement.",[12,3008,3009],{},"Second, for straightforward factual content — definitions, procedures, regulatory requirements, reference material — direct instruction is efficient and effective. The testing effect literature shows that retrieval practice helps retention, but the material being retrieved still needs to be clearly presented in the first place. No amount of interactivity compensates for unclear or disorganised content. A well-written reference document that people actually consult when they need it may do more than an interactive module they complete once and never revisit.",[12,3011,3012],{},"Third, there is the pragmatic argument. Interactive e-learning takes longer to design, longer to build, and longer to maintain. If the content changes frequently — as compliance regulations or product specifications often do — the maintenance burden of branching scenarios and embedded interactions can become unsustainable. A simpler format that gets updated promptly may serve learners better than an elaborate interactive module that sits unchanged for two years after a regulation changes.",[12,3014,3015],{},"These are not arguments against active learning. They are arguments for matching the design investment to the learning need.",[211,3017],{},[24,3019,3021],{"id":3020},"where-the-evidence-genuinely-conflicts","Where the evidence genuinely conflicts",[12,3023,3024],{},"There are areas where the research is less settled than advocates on either side tend to acknowledge.",[12,3026,3027,3030],{},[57,3028,3029],{},"Interactivity can hurt."," Several studies in multimedia learning have found that adding interactive elements to instructional materials can decrease learning outcomes compared to simpler formats. The mechanism is cognitive load: when the interaction itself is complex or unfamiliar, working memory resources that should be processing the content are instead consumed by figuring out the interface. This is the paradox at the heart of much e-learning: the features meant to increase engagement can reduce learning if they are not carefully designed. A drag-and-drop exercise that is mechanically fiddly, a branching scenario where the navigation is confusing, a simulation where the controls are unintuitive — all of these add extraneous cognitive load that competes with the instructional content.",[12,3032,3033,3036,3037,3042,3043,3046,3047,3049],{},[57,3034,3035],{},"Learners prefer what works less well."," The ",[171,3038,3041],{"href":3039,"rel":3040},"https:\u002F\u002Fwww.pnas.org\u002Fdoi\u002F10.1073\u002Fpnas.1821936116",[222],"Deslauriers et al. (2019)"," finding, mentioned in the first post of this series, bears repeating. Students who learned through active methods performed significantly better on assessments but reported ",[16,3044,3045],{},"lower"," satisfaction and ",[16,3048,3045],{}," perceived learning than students who attended polished lectures. Feelings of fluency — the sense that material is easy and clear — are not reliable indicators of durable learning. This creates a genuine tension for e-learning designers: the formats that produce the best feedback scores are often not the ones that produce the best retention. In organisations where course ratings drive decisions, this tension is structural.",[12,3051,3052,3055,3056,3061],{},[57,3053,3054],{},"Effect sizes are real but modest."," Even in the most favourable meta-analyses, the effect sizes for active learning strategies range from about 0.15 to 0.50 standard deviations. An effect size of 0.33 — the finding from a ",[171,3057,3060],{"href":3058,"rel":3059},"https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fabs\u002Fpii\u002FS1747938X25000454",[222],"recent 2025 meta-analysis on active strategies in video-based learning"," — means the average active learner performs at about the 63rd percentile of the passive group. This is meaningful, especially at scale. But it is not transformative. It does not mean that passive formats fail entirely or that active formats guarantee success. It means that, on average, across many studies, active approaches produce a moderate improvement. Whether that improvement justifies the additional development cost depends on context: the stakes of the learning, the size of the audience, and the longevity of the content.",[211,3063],{},[24,3065,3067],{"id":3066},"what-this-actually-means-for-design-decisions","What this actually means for design decisions",[12,3069,3070],{},"The research, taken as a whole, does not support either extreme. It does not support the claim that all e-learning should be interactive and scenario-based regardless of content or context. And it does not support the claim that passive delivery is just as effective and we should save the budget.",[12,3072,3073],{},"What it supports is a more honest set of design questions.",[12,3075,3076,3079],{},[57,3077,3078],{},"What kind of learning does this content require?"," If learners need to recall facts or follow procedures, direct instruction with spaced retrieval practice is efficient and well-supported. If learners need to exercise judgement, navigate ambiguity, or apply principles in unfamiliar situations, scenario-based and constructive approaches earn their development cost. Matching the method to the learning goal — not applying the same template to everything — is where the evidence points.",[12,3081,3082,3085],{},[57,3083,3084],{},"Is the interactivity doing cognitive work, or cosmetic work?"," A click-to-reveal panel and a retrieval question both count as \"interactive\" in an authoring tool. They are not equivalent. The question is whether the learner is being asked to think — to retrieve, to decide, to generate — or merely to perform a motor action to advance the content. If the interaction could be replaced by a page turn without any loss of cognitive engagement, it is not earning its place.",[12,3087,3088,3091],{},[57,3089,3090],{},"What is the realistic maintenance window?"," A branching scenario that is accurate and current teaches more than a slide deck. A branching scenario that has not been updated since the regulation changed teaches less than nothing. Design sophistication needs to be matched with realistic update commitments. For rapidly changing content, a simpler format that stays current may genuinely serve learners better.",[12,3093,3094,3097],{},[57,3095,3096],{},"What are the actual stakes of failure?"," For a compliance refresher where the real goal is documentation, the development investment in rich interactivity may not be justified. For training where poor performance has safety, financial, or reputational consequences — onboarding in regulated industries, customer-facing communication, clinical decision-making — the moderate effect size of active learning, applied across thousands of learners, translates to meaningful organisational impact.",[211,3099],{},[24,3101,1253],{"id":1252},[12,3103,3104],{},"Active learning works. The evidence is consistent, replicated across hundreds of studies, and the effect sizes are meaningful. Retrieval practice is one of the most reliably demonstrated findings in all of cognitive psychology. Scenario-based learning outperforms information delivery for applied skills. The ICAP hierarchy — Interactive over Constructive over Active over Passive — holds up across contexts.",[12,3106,3107],{},"But \"active learning works\" is not the same as \"every e-learning module should be interactive.\" The gains are moderate, not magical. Poorly designed interactivity can be worse than well-designed passive delivery. The development cost is real. And for some learning goals, straightforward direct instruction is the right answer.",[12,3109,3110],{},"The useful question is not whether active involvement is better in the abstract. It is where, specifically, the evidence says the extra investment produces returns that justify it — and where it does not. The research gives us enough to answer that question with reasonable precision, if we are willing to read all of it rather than just the parts that confirm what we were already planning to build.",[211,3112],{},[12,3114,3115],{},[16,3116,3117,3118,3122,3123,3122,3127,3122,3131,3122,3135,3122,3139,3122,3143,3122,3148,3122,3153,543],{},"Sources: ",[171,3119,3121],{"href":2912,"rel":3120},[222],"Clark (1983, 1994)","; ",[171,3124,3126],{"href":2927,"rel":3125},[222],"Kozma (1991, 1994)",[171,3128,3130],{"href":2974,"rel":3129},[222],"Freeman, Eddy, McDonough, Smith et al. (2014)",[171,3132,3134],{"href":2705,"rel":3133},[222],"Roediger & Karpicke (2006)",[171,3136,3138],{"href":2985,"rel":3137},[222],"Chi & Wylie (2014)",[171,3140,3142],{"href":3039,"rel":3141},[222],"Deslauriers, McCarty, Miller, Callaghan & Kestin (2019)",[171,3144,3147],{"href":3145,"rel":3146},"https:\u002F\u002Fwww.sri.com\u002Fwp-content\u002Fuploads\u002F2021\u002F12\u002Feffectiveness_of_online_and_blended_learning.pdf",[222],"Means, Toyama, Murphy, Bakia & Jones (2009\u002F2013)",[171,3149,3152],{"href":3150,"rel":3151},"https:\u002F\u002Fpsycnet.apa.org\u002Frecord\u002F2017-22562-006",[222],"Adesope, Trevisan & Sundararajan (2017)",[171,3154,3156,3157,3160],{"href":3058,"rel":3155},[222],"Meta-analysis on active learning strategies in video learning (2025, ",[16,3158,3159],{},"Learning and Instruction",")",{"title":226,"searchDepth":227,"depth":227,"links":3162},[3163,3164,3165,3166,3167,3168],{"id":2905,"depth":227,"text":2906},{"id":2949,"depth":227,"text":2950},{"id":2999,"depth":227,"text":3000},{"id":3020,"depth":227,"text":3021},{"id":3066,"depth":227,"text":3067},{"id":1252,"depth":227,"text":1253},"2026-03-29","The forty-year counter-tradition in learning research — Clark's media debate, the 'no significant difference' finding, and what they mean for e-learning.","\u002Fblog\u002Fcase-against-active-learning.webp",{},"\u002Fblog\u002Fthe-case-against-active-learning",{"title":2873,"description":3170},"blog\u002Fthe-case-against-active-learning",[249,250],"vo1_-guWXUhabt6L9K3jWKebgdZslEAhawUDDZ5XGvE",{"id":3179,"title":3180,"author":7,"body":3181,"date":3430,"description":3431,"extension":239,"image":3432,"meta":3433,"navigation":242,"path":3434,"seo":3435,"stem":3436,"tags":3437,"__hash__":3438},"blog\u002Fblog\u002Flearning-through-decision.md","Learning through decisions",{"type":9,"value":3182,"toc":3421},[3183,3188,3193,3195,3198,3201,3203,3207,3220,3223,3226,3228,3232,3235,3248,3251,3262,3269,3271,3275,3278,3285,3303,3306,3308,3312,3315,3322,3325,3328,3338,3344,3350,3352,3356,3359,3362,3365,3368,3370,3374,3377,3380,3383,3386,3389,3392,3395,3397,3401,3404,3407,3409,3414,3416],[3184,3185,3187],"h1",{"id":3186},"learning-through-decisions-what-the-research-says-about-scenario-based-design","Learning through decisions - what the research says about scenario-based design",[12,3189,3190],{},[16,3191,3192],{},"This is the second post in a series exploring the evidence behind learning design decisions. The first post examined why the tell-and-test format underperforms and what a century of retrieval practice research tells us about retention. This post builds on that foundation by looking at a different question: what happens when we ask learners to make decisions rather than absorb information?",[211,3194],{},[12,3196,3197],{},"In the previous post, we established that retrieval practice - being asked to recall or apply information - consistently outperforms re-reading and passive exposure. The testing effect is one of the most reliably replicated findings in educational psychology, with effect sizes around g = 0.50 across hundreds of studies.",[12,3199,3200],{},"But retrieval practice has a limit. Most of the evidence behind it involves recalling facts, concepts, and procedures - the kind of knowledge that can be tested with a question and a correct answer. Workplace learning often demands something different: the ability to read a situation, weigh competing considerations, and act appropriately when the textbook answer is not clearly labelled. That is where scenario-based learning enters - and where the research gets genuinely interesting.",[211,3202],{},[24,3204,3206],{"id":3205},"the-problem-with-knowing-without-being-able-to-use","The problem with knowing without being able to use",[12,3208,3209,3210,3213,3214,728,3216,3219],{},"There is a well-documented gap in learning research between declarative knowledge - knowing ",[16,3211,3212],{},"that"," something is true - and procedural or applied knowledge - knowing ",[16,3215,1775],{},[16,3217,3218],{},"when"," to use it. The gap shows up most clearly when learners who can answer factual questions correctly still fail to apply the underlying principles in unfamiliar situations.",[12,3221,3222],{},"This is not a failure of motivation or attention. It reflects something fundamental about how knowledge is stored and retrieved. Information learned in one context - a course, a slide, a definition - is not automatically available in a different context, like a conversation with a difficult client or an ambiguous ethical situation at work. Cognitive psychologists call this the transfer problem, and it is the central challenge of training design.",[12,3224,3225],{},"Scenario-based learning addresses the transfer problem directly. Rather than teaching principles and then hoping learners will apply them, it places learners inside a situation that demands application. The sequence is reversed: encounter the problem first, work through it, then surface the principle. The learning is anchored to a context that resembles the one where it will actually be needed.",[211,3227],{},[24,3229,3231],{"id":3230},"what-the-meta-analyses-actually-show","What the meta-analyses actually show",[12,3233,3234],{},"The research on problem-based and scenario-based learning is broadly positive but contains important nuance that is worth understanding before reaching for it as a design default.",[12,3236,3237,3238,3240,3241,3244,3245,543],{},"Dochy, Segers, Van den Bossche, and Gijbels (2003, ",[16,3239,3159],{},") conducted a meta-analysis of 43 studies comparing problem-based learning with conventional instruction. Their summary has become one of the most quoted - and most misquoted - findings in instructional design: students in problem-based conditions showed ",[57,3242,3243],{},"robust positive effects on skills and knowledge application",", with no single study reporting a negative result, while showing a ",[57,3246,3247],{},"slight negative effect on factual knowledge recall",[12,3249,3250],{},"The misquotation usually stops at the negative finding. The full picture is more interesting: students in problem-based conditions gained slightly less knowledge in the short term, but retained significantly more of what they did gain over time. As Dochy and colleagues put it, they \"gained slightly less knowledge, but remember more of the acquired knowledge.\"",[12,3252,3253,3254,3257,3258,3261],{},"Strobel and van Barneveld (2009) synthesised eight prior meta-analyses and confirmed the pattern. Problem-based learning was ",[57,3255,3256],{},"superior for long-term retention, skill development, and learner satisfaction",". Traditional instruction was ",[57,3259,3260],{},"superior for short-term retention on standardised tests",". If your success metric is a knowledge check at the end of a course, conventional instruction may win. If your metric is behaviour six months later, the picture reverses.",[12,3263,3264,3265,3268],{},"Gijbels, Dochy, Van den Bossche, and Segers (2005) added an important refinement: the advantage of problem-based approaches grew largest when assessments measured ",[57,3266,3267],{},"understanding of principles linking concepts"," rather than recall of isolated facts. This is the kind of understanding that actually predicts performance in complex roles.",[211,3270],{},[24,3272,3274],{"id":3273},"the-productive-failure-finding","The productive failure finding",[12,3276,3277],{},"The most counterintuitive - and practically useful - finding in this space comes from Manu Kapur's research programme on what he calls productive failure.",[12,3279,3280,3281,3284],{},"The central claim is this: asking learners to attempt a problem ",[16,3282,3283],{},"before"," receiving instruction produces better learning than providing instruction first and then asking learners to practise. Not because they solve the problem correctly - most do not. But because the struggle activates prior knowledge, surfaces the gaps in existing understanding, and creates a kind of cognitive readiness that makes the subsequent instruction far more effective.",[12,3286,3287,3288,3290,3291,3294,3295,3298,3299,3302],{},"Sinha and Kapur's (2021, ",[16,3289,2917],{},") meta-analysis synthesised 53 studies involving over 12,000 participants. Problem-solving before instruction outperformed instruction before problem-solving at ",[57,3292,3293],{},"g = 0.36"," overall, rising to ",[57,3296,3297],{},"g = 0.58"," with high-fidelity implementation of the productive failure design. After adjusting for publication bias, the effect for conceptual knowledge and transfer reached ",[57,3300,3301],{},"g = 0.87",". Crucially, productive failure produced no compromise on procedural knowledge - learners matched direct instruction on procedural tasks while substantially exceeding them on conceptual understanding and transfer.",[12,3304,3305],{},"For scenario-based e-learning, this has a direct design implication. The instinct is usually to front-load information - teach the concept, then present the scenario as practice. The evidence suggests inverting this, at least for conceptual learning: present the scenario first, let learners attempt it with what they already know, then use the debrief to introduce the principle. The difficulty is not a design flaw. It is the mechanism.",[211,3307],{},[24,3309,3311],{"id":3310},"what-makes-a-scenario-actually-work","What makes a scenario actually work",[12,3313,3314],{},"Not all scenarios are equal, and the research helps explain why some produce genuine learning while others feel like window dressing on a multiple-choice question.",[12,3316,3317,3318,3321],{},"The critical variable is ",[57,3319,3320],{},"decision authenticity",". Scenarios that present obviously correct answers alongside clearly wrong distractors require recognition, not reasoning. They test whether learners can identify the right answer when it is labelled by context - which is a very different skill from navigating a genuinely ambiguous situation where reasonable people might disagree.",[12,3323,3324],{},"Chi and Wylie's ICAP framework, discussed in the previous post, is useful here. A scenario that asks learners to choose between \"report the issue immediately\" and \"ignore it and hope it resolves\" is operating in the Active mode at best - learners are doing something, but not necessarily thinking. A scenario that presents a situation where reporting has costs, not reporting has risks, the right authority is unclear, and time pressure is real - that forces Constructive engagement. Learners must generate a position, not recognise one.",[12,3326,3327],{},"Several design features are supported by evidence as genuinely meaningful rather than cosmetic.",[12,3329,3330,3333,3334,3337],{},[57,3331,3332],{},"Consequences matter more than correctness."," Evans and Gibbons (2007, ",[16,3335,3336],{},"Computers & Education",") found that interactive e-learning significantly improved transfer and problem-solving but showed no significant advantage on factual recall. The mechanism was consequential feedback - seeing what their decision caused - rather than being told they were right or wrong. Consequences create the emotional and cognitive trace that makes the scenario memorable and applicable. A wrong answer followed by a natural consequence teaches more than a wrong answer followed by \"incorrect - try again.\"",[12,3339,3340,3343],{},[57,3341,3342],{},"Branching adds value when it reflects real complexity."," Branching scenarios that allow different paths to produce genuinely different outcomes reflect the actual structure of complex decisions. Branching that reconverges after one or two steps - where the story resets regardless of what the learner chose - provides the appearance of consequence without the substance. Learners notice.",[12,3345,3346,3349],{},[57,3347,3348],{},"Characters and context reduce abstraction."," There is consistent evidence that grounding problems in realistic characters and contexts aids transfer compared to abstract case descriptions. This is one mechanism behind the spacing and interleaving effects: context-rich scenarios help learners build flexible representations of when and where to apply a principle, rather than a single rigid association between a cue and an answer.",[211,3351],{},[24,3353,3355],{"id":3354},"the-important-counterargument","The important counterargument",[12,3357,3358],{},"In 2006, Kirschner, Sweller, and Clark published an influential paper arguing that minimally guided instruction - which includes discovery learning and many constructivist approaches - consistently underperforms explicit instruction for novice learners. The paper has over 4,000 citations and is regularly used to argue against scenario-based approaches.",[12,3360,3361],{},"The argument deserves engagement rather than dismissal. Cognitive load theory, which underpins it, is well-supported. Novice learners lack the schemas that allow them to distinguish relevant from irrelevant information in a complex scenario. Without those schemas, the cognitive load of navigating an unfamiliar situation can overwhelm working memory, leaving little capacity for actual learning. The result is busy-feeling activity with limited retention.",[12,3363,3364],{},"The productive resolution, developed by Hmelo-Silver, Duncan, and Chinn (2007) and supported by the productive failure evidence, is that the dichotomy between guided and unguided instruction is false. Well-designed scenario-based learning is not minimally guided - it is specifically guided. The scaffolding is embedded in the scenario structure itself: the situation is complex but bounded, the decision points are clear, the consequences are informative, and the debrief provides the explicit instruction that connects the experience to the underlying principle.",[12,3366,3367],{},"The expertise reversal effect adds a further refinement: the amount of scaffolding needed decreases as expertise grows. A scenario that appropriately challenges a new compliance officer will bore a senior one. Designing for a single learner profile is always a compromise - and the cost of that compromise falls on whoever is furthest from the assumed baseline.",[211,3369],{},[24,3371,3373],{"id":3372},"what-this-suggests-about-design","What this suggests about design",[12,3375,3376],{},"The research does not suggest that every piece of e-learning should be a scenario. Factual knowledge - definitions, procedures, regulatory requirements - can be taught efficiently through direct instruction, and adding a scenario wrapper adds development time without necessarily improving outcomes for straightforward content.",[12,3378,3379],{},"Where scenario-based learning earns its complexity is exactly the territory where tell-and-test fails: situations requiring judgement, situations where the stakes of a wrong decision are real, situations where the goal is not just remembering a rule but knowing when and how to apply it under pressure.",[12,3381,3382],{},"A few principles follow from the evidence.",[12,3384,3385],{},"Present the problem before the principle, at least some of the time. Let learners struggle with a situation using what they already know, then use the debrief to introduce or reinforce the underlying concept. The struggle is not a problem to be designed away - it is doing some of the learning.",[12,3387,3388],{},"Design for authentic decisions, not recognisable answers. If a reasonably attentive learner can identify the correct path without engaging with the substance of the scenario, the scenario is not doing its job. The decision should require genuine reasoning about the specific situation, not pattern-matching to \"what does a compliance course want me to say?\"",[12,3390,3391],{},"Use consequences rather than judgements as feedback. Showing learners what their decision causes - in the logic of the scenario - teaches more than telling them they were right or wrong. The consequence anchors the learning to the decision rather than to the assessment.",[12,3393,3394],{},"And match scenario complexity to learner expertise. What challenges a novice overwhelms a beginner and bores an expert. A single scenario designed for the average learner is serving no one especially well.",[211,3396],{},[24,3398,3400],{"id":3399},"a-note-on-ai-and-scenarios","A note on AI and scenarios",[12,3402,3403],{},"One practical constraint on scenario-based learning has always been development time. A branching scenario with multiple paths, realistic characters, and meaningful consequences takes considerably longer to build than a slide-and-test format. This is a real cost, and for many organisations it is the reason tell-and-test persists despite its limitations - not ignorance of the research, but production economics.",[12,3405,3406],{},"AI-assisted authoring changes that calculation in ways that are still being worked out. Generating plausible character dialogue, drafting decision branches, and creating consequence text are exactly the kinds of tasks where AI can accelerate production without compromising the instructional logic - provided the instructional logic is supplied by the designer. The research on what makes scenarios effective is not something AI can substitute for. But the time cost of building out scenarios once the logic is established is substantially lower than it was five years ago.",[211,3408],{},[12,3410,3411],{},[16,3412,3413],{},"The next post in this series will examine a finding that most e-learning design ignores entirely: the expertise reversal effect, and what it means for courses designed for mixed-experience audiences.",[211,3415],{},[12,3417,3418],{},[16,3419,3420],{},"Sources: Dochy, Segers, Van den Bossche & Gijbels (2003); Strobel & van Barneveld (2009); Gijbels, Dochy, Van den Bossche & Segers (2005); Sinha & Kapur (2021); Chi & Wylie (2014); Evans & Gibbons (2007); Kirschner, Sweller & Clark (2006); Hmelo-Silver, Duncan & Chinn (2007).",{"title":226,"searchDepth":227,"depth":227,"links":3422},[3423,3424,3425,3426,3427,3428,3429],{"id":3205,"depth":227,"text":3206},{"id":3230,"depth":227,"text":3231},{"id":3273,"depth":227,"text":3274},{"id":3310,"depth":227,"text":3311},{"id":3354,"depth":227,"text":3355},{"id":3372,"depth":227,"text":3373},{"id":3399,"depth":227,"text":3400},"2026-03-17","There is a well-documented gap in learning research between declarative knowledge - knowing *that* something is true - and procedural or applied knowledge - knowing *how* and *when* to use it. The gap shows up most clearly when learners who can answer factual questions correctly still fail to apply the underlying principles in unfamiliar situations.","\u002Fblog\u002Flearning-through-decisions.webp",{},"\u002Fblog\u002Flearning-through-decision",{"title":3180,"description":3431},"blog\u002Flearning-through-decision",[249,250],"gHxvWxgrGRfZW3hUXa08i0AfyvL57JsSe8hQvDKwu9E",{"id":3440,"title":3441,"author":7,"body":3442,"date":3662,"description":3450,"extension":239,"image":3663,"meta":3664,"navigation":242,"path":3665,"seo":3666,"stem":3667,"tags":3668,"__hash__":3669},"blog\u002Fblog\u002Fwhy-most-e-learning-fails.md","Why most e-learning doesn't work",{"type":9,"value":3443,"toc":3654},[3444,3448,3451,3454,3456,3460,3463,3466,3468,3472,3475,3491,3500,3509,3511,3515,3518,3525,3528,3544,3547,3549,3553,3556,3566,3592,3599,3602,3605,3607,3611,3614,3617,3626,3634,3636,3638,3641,3644,3647,3649],[3184,3445,3447],{"id":3446},"why-most-e-learning-doesnt-work-and-what-the-research-actually-says","Why most e-learning doesn't work - and what the research actually says",[12,3449,3450],{},"There is a moment most instructional designers recognise. You finish a course. It looks good. The client approves it. Learners click through, hit the knowledge check at the end, score 80%, and receive their completion certificate. Six weeks later, you wonder whether any of it stuck.",[12,3452,3453],{},"The uncomfortable answer, if you follow the research, is probably not much.",[211,3455],{},[24,3457,3459],{"id":3458},"the-tell-and-test-problem","The tell-and-test problem",[12,3461,3462],{},"The dominant format in organisational e-learning follows a simple pattern: present information, then test whether learners absorbed it. Slides of content, perhaps with a narrator, followed by a handful of multiple-choice questions to confirm completion. This format is efficient to produce, easy to track, and almost universally used. It is also, by the standards of learning science, one of the least effective approaches available.",[12,3464,3465],{},"The issue is not that testing is bad. Testing is actually one of the most powerful tools we have. The issue is the sequence: present everything first, test at the very end. By that point, the test is measuring short-term recognition of recently seen material, not durable learning. Pass rates look good. Retention curves tell a different story.",[211,3467],{},[24,3469,3471],{"id":3470},"what-a-century-of-research-actually-shows","What a century of research actually shows",[12,3473,3474],{},"The evidence that active learning outperforms passive delivery is not new, not thin, and not limited to schools or universities.",[12,3476,3477,3478,3486,3487,3490],{},"A landmark ",[171,3479,3482,3483],{"href":3480,"rel":3481},"https:\u002F\u002Fwww.pnas.org\u002Fdoi\u002F10.1073\u002Fpnas.1319030111",[222],"2014 meta-analysis published in ",[16,3484,3485],{},"PNAS"," analysed 225 studies across undergraduate science, engineering and mathematics. Students in active learning conditions scored an average of nearly half a standard deviation higher on exams. Students in lecture-only formats were ",[57,3488,3489],{},"1.5 times more likely to fail",". The researchers described continuing to teach by lecture alone, given this evidence, as a kind of experiment on students - one without their informed consent.",[12,3492,3493,3494,3499],{},"In 2011, a ",[171,3495,3498],{"href":3496,"rel":3497},"https:\u002F\u002Fpubmed.ncbi.nlm.nih.gov\u002F21566198\u002F",[222],"controlled experiment at the University of British Columbia"," offered an even sharper result. Researchers compared an experienced, highly-rated lecturer against an instructor with no teaching experience using active methods. The active learning group outperformed the experienced lecturer by more than double on an identical assessment - 74% versus 41%. The medium did not determine the outcome. The method did.",[12,3501,3502,3503,3508],{},"These are not isolated findings. Effect sizes around d = 0.47 appear consistently across subjects and settings, including a ",[171,3504,3507],{"href":3505,"rel":3506},"https:\u002F\u002Flink.springer.com\u002Farticle\u002F10.1007\u002Fs10734-022-00977-8",[222],"2022 meta-analysis"," extending the evidence to humanities and social sciences. The direction of the effect is one of the most reliably replicated findings in educational research.",[211,3510],{},[24,3512,3514],{"id":3513},"the-testing-effect-the-most-underused-tool-in-e-learning","The testing effect: the most underused tool in e-learning",[12,3516,3517],{},"If you had to choose a single finding from cognitive psychology to build a learning design philosophy around, it would probably be the testing effect.",[12,3519,3520,3521,3524],{},"In a now-classic series of experiments, ",[171,3522,2707],{"href":2705,"rel":3523},[222]," asked students to study a passage of text using two different approaches. One group read it four times. Another read it once, then recalled as much as they could - three times, without looking at the material. On an immediate test, the re-reading group performed slightly better. One week later, the retrieval group outperformed them by 21 percentage points. Restudying led to 56% forgetting over two days. Retrieval practice produced only 13%.",[12,3526,3527],{},"The mechanism matters here. When you retrieve something from memory, you are not just checking whether it is there - you are strengthening the memory itself. Each retrieval makes the next retrieval easier and more durable. Re-reading, by contrast, creates a fluency illusion: the material feels familiar, which learners mistake for knowing it.",[12,3529,3530,3531,3535,3536,3539,3540,3543],{},"A meta-analysis by ",[171,3532,3534],{"href":3150,"rel":3533},[222],"Adesope, Trevisan, and Sundararajan (2017)"," synthesised 272 effects from 118 articles. The advantage of retrieval practice over restudying: ",[57,3537,3538],{},"g = 0.51",". Over no-activity controls: ",[57,3541,3542],{},"g = 0.93",". The effect held in classrooms as well as laboratories.",[12,3545,3546],{},"For e-learning designers, this has a direct implication. Questions are not just assessment tools. Placed before content, they prime attention. Placed during content, they interrupt passive processing at exactly the right moment. Placed after content - especially after a delay - they do more for long-term retention than any amount of re-exposure to the material. The tell-and-test format gets the sequence backwards.",[211,3548],{},[24,3550,3552],{"id":3551},"not-all-clicks-are-equal","Not all clicks are equal",[12,3554,3555],{},"It is tempting to conclude from this that interactive e-learning is simply better than non-interactive. The reality is more specific - and more useful.",[12,3557,3558,3565],{},[171,3559,3561,3562,3160],{"href":2985,"rel":3560},[222],"Michelene Chi and Ruth Wylie's ICAP framework (2014, ",[16,3563,3564],{},"Educational Psychologist"," offers a practical taxonomy. They distinguish four modes of engagement:",[455,3567,3568,3574,3580,3586],{},[458,3569,3570,3573],{},[57,3571,3572],{},"Interactive"," - co-constructing through dialogue or debate",[458,3575,3576,3579],{},[57,3577,3578],{},"Constructive"," - generating new output: self-explanations, decisions, predictions",[458,3581,3582,3585],{},[57,3583,3584],{},"Active"," - manipulating materials: clicking, highlighting, dragging",[458,3587,3588,3591],{},[57,3589,3590],{},"Passive"," - receiving without overt response",[12,3593,3594,3595,3598],{},"The predicted learning outcomes follow in that order, with one critical detail: the largest jump in learning occurs ",[57,3596,3597],{},"between Active and Constructive",". Much of what e-learning calls \"interactive\" - click-to-reveal panels, drag-and-drop sorting, navigation controls - falls into the merely Active category. Learners are doing something, but not necessarily thinking.",[12,3600,3601],{},"This matters because there is consistent evidence that high interactivity without cognitive purpose can actually harm learning. Extraneous interactions consume working memory resources that would otherwise be used to process the material itself. A study in medical education found that simpler \"click\" interactions outperformed more complex \"drag\" interactions precisely because the cognitive load of the manipulation competed with content understanding. Several studies now document an inverted-U relationship between interactivity and learning outcomes: moderate interactivity tends to produce the best results.",[12,3603,3604],{},"The implication is not to make e-learning simpler. It is to make interactivity earn its place. A well-designed decision point in a scenario, where learners must commit to a choice and live with the consequences, generates far more learning than ten click-to-reveal panels. A self-explanation prompt - \"why do you think that approach would work here?\" - activates constructive processing that passive delivery never reaches.",[211,3606],{},[24,3608,3610],{"id":3609},"an-honest-caveat","An honest caveat",[12,3612,3613],{},"The research is clear in direction but comes with genuine limitations worth naming.",[12,3615,3616],{},"Most studies measuring active learning advantages use short-term retention in academic settings with motivated student populations. The effect sizes in corporate, self-paced, optional e-learning - where attention is divided, stakes feel lower, and completion is often the actual goal - may be smaller. Large-scale randomised trials in workplace learning contexts remain scarce.",[12,3618,3619,3620,3625],{},"There is also the expertise reversal effect, documented across dozens of ",[171,3621,3624],{"href":3622,"rel":3623},"https:\u002F\u002Fwww.researchgate.net\u002Fpublication\u002F48829036_The_Expertise_Reversal_Effect",[222],"studies by Kalyuga, Sweller, and colleagues",": instructional scaffolding that helps novices can become redundant or even counterproductive for experienced learners. A compliance course that walks a new hire through every principle step by step will frustrate a seasoned professional. The same interactivity level for all learners is always a compromise.",[12,3627,3628,3629,3633],{},"And perhaps most striking: a ",[171,3630,3632],{"href":3039,"rel":3631},[222],"Harvard crossover study (Deslauriers et al., 2019)"," found that students consistently preferred passive lectures but learned significantly more from active learning. Learner satisfaction scores and actual learning outcomes were negatively correlated. This should give anyone designing for learner feedback or course ratings some pause - the formats that feel easiest often teach least.",[211,3635],{},[24,3637,3373],{"id":3372},[12,3639,3640],{},"The evidence points less toward a specific format and more toward a few persistent principles.",[12,3642,3643],{},"Make learners retrieve before you tell them the answer. Space questions across a course rather than bunching them at the end. Use scenarios that require genuine decisions, not just recognition of the obvious wrong answer. Match complexity to learner expertise. And be honest with clients about what completion metrics measure - and what they do not.",[12,3645,3646],{},"This is not a new insight. Learning designers have been making these arguments for decades. The research simply gives those arguments firmer ground.",[211,3648],{},[12,3650,3651],{},[16,3652,3653],{},"This is the first in a series exploring the evidence behind learning design decisions - what the research actually supports, where it is mixed, and how it shapes the way we build courses at LearnBuilder.",{"title":226,"searchDepth":227,"depth":227,"links":3655},[3656,3657,3658,3659,3660,3661],{"id":3458,"depth":227,"text":3459},{"id":3470,"depth":227,"text":3471},{"id":3513,"depth":227,"text":3514},{"id":3551,"depth":227,"text":3552},{"id":3609,"depth":227,"text":3610},{"id":3372,"depth":227,"text":3373},"2026-03-14","\u002Fblog\u002Fwhy-most-e-learning-does-not-work.webp",{},"\u002Fblog\u002Fwhy-most-e-learning-fails",{"title":3441,"description":3450},"blog\u002Fwhy-most-e-learning-fails",[249,250],"mrvAqr0Hl9lMXRfYFYXAguylncz28bTDyyNnYGZ7i38",{"id":3671,"title":3672,"author":7,"body":3673,"date":3734,"description":3735,"extension":239,"image":3736,"meta":3737,"navigation":242,"path":3738,"seo":3739,"stem":3740,"tags":3741,"__hash__":3744},"blog\u002Fblog\u002Fwelcome-to-learnbuilder.md","Welcome to the LearnBuilder Blog",{"type":9,"value":3674,"toc":3729},[3675,3678,3682,3685,3711,3715,3718,3722],[12,3676,3677],{},"We're excited to launch the LearnBuilder blog! This is where we'll share our thoughts on e-learning, instructional design, and how AI is changing the way we create courses.",[24,3679,3681],{"id":3680},"what-to-expect","What to expect",[12,3683,3684],{},"We'll be writing about:",[455,3686,3687,3693,3699,3705],{},[458,3688,3689,3692],{},[57,3690,3691],{},"Product updates"," - new features, improvements, and what's coming next",[458,3694,3695,3698],{},[57,3696,3697],{},"Instructional design"," - practical tips for creating engaging and meaningful learning experiences",[458,3700,3701,3704],{},[57,3702,3703],{},"AI in e-learning"," - how we use AI to make course creation faster and better, and where the industry is heading",[458,3706,3707,3710],{},[57,3708,3709],{},"Case studies"," - real examples of what people are building with LearnBuilder",[24,3712,3714],{"id":3713},"why-we-started-learnbuilder","Why we started LearnBuilder",[12,3716,3717],{},"After 15 years of designing learning experiences for thousands of learners across 150 countries, one thing became clear: the tools available were either too slow, too expensive, or too boring. LearnBuilder was built to change that - combining real instructional design thinking with the speed AI now makes possible.",[24,3719,3721],{"id":3720},"stay-tuned","Stay tuned",[12,3723,3724,3725,543],{},"We have a lot of exciting content planned. If you have topics you'd like us to cover, don't hesitate to reach out at ",[171,3726,3728],{"href":3727},"mailto:contact@learnbuilder.org","contact@learnbuilder.org",{"title":226,"searchDepth":227,"depth":227,"links":3730},[3731,3732,3733],{"id":3680,"depth":227,"text":3681},{"id":3713,"depth":227,"text":3714},{"id":3720,"depth":227,"text":3721},"2026-03-13","Introducing the LearnBuilder blog - where we share insights on AI-powered course creation, instructional design tips, and product updates.","\u002Fblog\u002FWelcome-to-the-LearnBuilder-Blog.webp",{},"\u002Fblog\u002Fwelcome-to-learnbuilder",{"title":3672,"description":3735},"blog\u002Fwelcome-to-learnbuilder",[3742,3743],"announcement","product","nBxDUQo6MxKsGvM3YvUKtpKEjIgdAs31L2rXfCgFxvs",1779992475039]