Content Personalization for Learners

I'm guilty of saying vague things like "AI helps us personalize learning", but we should get more specific. Here's a better framework: **Dimension 1: Personalize TO** - Persona (role, demographics, interest groups) - Individual (learner history, goals, preferences, skills, achievements) - Context (environment, situation, current activity/task, external conditions) - Dynamic Adaptation (real-time behaviors, emotional/cognitive state, immediate interactions) **Dimension 2: Personalize WITH** - Content & Resources (examples, scenarios, multimedia, exercises tailored to learner) - Instructional Strategies (methods such as scaffolding, exploratory learning, collaborative vs. individual tasks) - Pacing & Sequencing (rate of instruction, order of activities/modules, complexity adjustment) - Assessment & Feedback (adaptive quizzes, diagnostic evaluations, targeted formative feedback) - Motivational Elements (gamification, goal-setting, rewards, incentives, personalized recognition) - Interface & Interaction (UX design, modality—visual/audio/tactile, navigation paths, accessibility customizations) **Dimension 3: Personalization PURPOSE** - Engagement & Motivation (increase learner interest, attention, enjoyment, participation) - Performance Improvement (enhance learner outcomes, skills development, mastery) - Accessibility & Inclusion (address diverse learner needs, equity, remove barriers) - Efficiency & Time Optimization (reduce learning time, improve instructional efficiency, avoid redundancy) - Knowledge Retention & Transfer (long-term retention, real-world application, deeper understanding) We shouldn't fall for generic AI hype.... this type of framework can help us be specific about what we mean by personalization.

For the past decade many #learning organizations dreamt of adopting the Netflix model. But it was just a dream and frankly, that model would have been a bad-fit in the corporate learning space anyway. However, something interesting just happened that I believe tech minded Learning Leaders / LXM, LXP solution providers should pay attention to. Netflix just published a really interesting Tech Blog article on their foundation model for personalized recommendations. It offers valuable insights that can inform the redesign of learner experiences. Key Takeaways: Centralized Learner Modeling: Netflix transitioned from multiple specialized models to a unified foundation model, centralizing user preference learning. * Application in Learning: Develop a centralized learner model that aggregates data from various learning activities, enabling consistent personalization across different courses and modules. Data-Centric Approach: Emphasizing high-quality, large-scale data over intricate feature engineering, Netflix’s model benefits from end-to-end learning. * App. In Lrg: Prioritize collecting comprehensive learner interaction data (e.g., quiz attempts, forum participation) to inform adaptive learning paths and content recommendations. Interaction Tokenization: Netflix tokenizes user interactions to capture meaningful sequences, similar to language models. * App. In Lrg: Implement tokenization of learning activities to identify patterns (e.g., common misconceptions, preferred learning sequences) that can guide personalized content delivery. Scalable Personalization: The foundation model allows for scalable personalization across Netflix’s vast user base. * App. In Lrg: Design learning systems that can scale personalization efforts, accommodating diverse learner profiles and adapting to evolving educational needs. Interaction tokenization (IT) sounded very similar to LRS (Lrg Record Store). IT is the process of converting user activities (e.g., watching a video, taking a quiz, clicking “next,” participating in a forum, pausing content, revisiting materials) into “tokens”—discrete data units. These tokens form sequences that can be analyzed like language to model and predict learner behavior or preferences. It’s like treating a learner’s journey as a sentence, where: Each “word” is an interaction (e.g., “view_video,” “attempt_quiz,” “fail_question_2”). The “sentence” is a learning path. The model learns from many such “sentences” to predict and personalize future experiences. LRS is the source: It captures and stores granular learning data in a structured format using xAPI statements. Tokenization is the next layer: Once you have data in the LRS, tokenization transforms these raw interactions into meaningful sequences for: Personalization Predictive analytics Content recommendation Learner path modeling (like Netflix does) Really interesting stuff. Give the article a read- link in comments.

Today, I would like to share an AI SoTL article entitled, “Experimentally testing AI-powered content transformations on student learning” by Heldreth et al. (2025) (https://lnkd.in/eanRDerM ). This study provides evidence that AI can measurably improve student learning outcomes when used to transform academic content. In a between-subjects experimental design with 60 U.S. high-school students, researchers compared learning a neuroscience textbook chapter using either a traditional digital PDF reader or an AI-powered platform called Learn Your Way, which transformed the same content into multiple interactive formats (immersive text, quizzes, slides, audio lessons, videos, and mind maps). The results were consistent and statistically significant. Students using the AI-powered system demonstrated higher immediate recall and superior long-term retention (3–7 days later) compared to those using the digital reader. Importantly, performance gains were not attributable to differences in prior knowledge, reading ability, interest, or assessment difficulty all were carefully controlled. Beyond test scores, students using Learn Your Way reported more positive learning experiences, including greater perceived understanding, higher enjoyment, stronger confidence, and a greater desire to reuse the tool. Qualitative data revealed why: students valued multimodal representations, chunked content, embedded quizzes, and timely feedback, all of which supported metacognitive monitoring and reduced cognitive overload. Grounded in multimedia learning theory, dual-coding theory, and self-directed learning principles, this study reinforces that AI is most effective when it re-represents content in cognitively supportive ways, rather than simply generating answers. Notably, learning gains were driven less by the number of AI features used and more by student agency in choosing representations that matched their learning needs. For teaching and learning, the implication is that AI can be used as a learning architecture, one that supports retrieval practice, feedback, personalization, and learner control at scale. Reference Heldreth, C., Vardoulakis, L. M., Miller, N. E., Haramaty, Y., Akrong, D., Hackmon, L., & Belinsky, L. (2025). Experimentally testing AI-powered content transformations on student learning. arXiv.

AI isn’t here to replace IDs. It’s here to amplify what we do best. I recently came across a study by Dr. Tian Luo et al. on instructional designers’ experience with generative AI. Turns out, most of us are already embracing it: for brainstorming, low-stakes proofreading, design support, and stakeholder collaboration. That aligns with what we’ve seen at ID Mentors. AI streamlines repetitive tasks, frees up creative space, and lets us focus on strategy and learner experience. Here’s how I see AI shifting the field, and how IDs can lead confidently in this new era: 1. AI-as-assistant, not autopilot Tools like ChatGPT or specialized assistants (Khan Academy, Kira Learning) are becoming like having a few graduate interns on standby. But—as educators rightly stress—human oversight remains essential. AI suggestions still need your instructional judgment to stay ethical, accurate, and learner-centered. 2. Ideation & prototyping, supercharged Need new scenarios, quizzes, or discussion prompts? AI generates solid first drafts instantly. So you can test ideas fast, iterate faster, and skip blank-slide paralysis. 3. Personalization at scale Forget one-size-fits-all. AI allows for adaptive pathways—quizzes that adjust difficulty, feedback that meets learners in real-time, materials that respond to individual performance. 4. Performance insights for real impact AI analytics can flag struggling learners, predict trend changes, even suggest resource adaptations—before issues spiral . That’s intelligent support at scale. But there are caveats. Let this guide how you use AI: Ethics, transparency & data privacy must stay central. We cannot outsource our responsibility to verify, credit, and ensure fairness. Human insight is irreplaceable. AI can assist—but it can’t reason, empathize, or connect with human nuance. This is a watershed moment for IDs. When we blend AI’s speed with our human purpose, we create smarter, more relevant, more responsive learning. What are your thoughts about how AI will influence ID? Let me know in the comments! #instructionaldesign #AIinLearning #futureofwork #IDMentors #theIndianID #learningdesign

The one-size-fits-all model just doesn’t work in education, and I admit that most EdTech companies (including us at Airtribe) are not solving this problem. In my experience, true learning is driven by curiosity, and that curiosity varies from person to person. The learning path isn’t linear for everyone. Some prefer diving deep like a DFS, exploring every detail in-depth, while others prefer a broad overview like a BFS, covering multiple concepts quickly to get the bigger picture. At Airtribe, while we offer extensive knowledge transfer through live sessions, we realized this is super useful but isn’t the most effective approach for every learner because everyone has a different starting point. So, we started exploring how to make learning more personalized, and Generative AI emerged as the perfect solution. Over the past few months, we’ve developed features to enhance the learning experience. One of the major additions is interactive reading components — a blend of text, code, videos, and quizzes designed to create a more engaging learning environment. But the 10x improvement is our new AI-driven nudges. These nudges prompt learners to explore more about a topic in a way that suits their learning style. If you’re curious, the AI will guide you to dive deeper and learn in a way that feels natural to you. We’re currently testing this with a small cohort, and the results are looking great. It's still early, but I believe this will significantly improve the way people learn on our platform. — Here’s an example of how someone (like me who prefers more examples) can learn about North Star Metrics while going through the reading content. 👇🏻

The people buying your course are not paying for access to your brilliance. They’re paying for a result. A real, visible shift in behavior that drives business outcomes. Unfortunately, most platforms weren’t built to support real behavior change. They were built to make it easier for you as the expert to record a video, hit publish, and collect payment. So, if you're choosing a platform based on how fast it helps you launch… …but not how well it helps your learners change… You’re setting yourself (and them) up for failure. After reviewing 30+ course platforms this year (like, for real), I’ve identified 9 key platform features that support real behavior change. 🧠 1. Cues & cadence Learners log in. Binge a few videos. Vanish. That’s a platform problem. Look for tools that let you pace the learning, drip content over time, and send timely nudges. 🪜 2. Ability scaffolding Throwing people into the deep end doesn’t build skills. It builds overwhelm. You want guardrails. Ways to unlock content after learners show mastery. ⚡ 3. Motivation mechanics You’re not just competing with other courses. You’re competing with Netflix. Points, badges, progress bars, when done well, help learners stay engaged, even when their motivation dips. 👥 4. Social accountability We’re herd animals and learning alongside others feels safe. Community spaces, peer check-ins, public progress sharing...all of these features boost engagement and follow-through. 🎯 5. Personalization Not every learner needs the same path. Select a platform that lets you route learners based on job role, existing skill level, or learning goals. 🧘 6. Reflection & practice Watching content ≠ behavior change. Look for platforms that support reflection exercises including: journal prompts, open-ended responses, or peer feedback forums. 📊 7. Progress visibility Tiny dopamine hits from seeing progress? They work. (Just ask Amazon why they added the fireworks emoji when you put something in your shopping cart). Progress bars, dashboards, and checklists hook learners so they keep going. 🌍 8. Environmental fit Is this a cohort? Self-paced? Blended? The right platform fits your course model, not the other way around. 🔁 9. Sustained support Habits don’t stick after 8 weeks of content. Look for features that help you deliver nudges, refreshers, and next steps content long after the course ends. Behavior change isn’t about what you teach. It’s about what your audience does with it...repeatedly. So if your course is promising outcomes… …your platform needs to be built for them. Curious where your current setup might be falling short? Send me a DM and we can set up a complimentary course audit for you. 👉 Follow Erin Green for tools on behavior change and course design. 🔁 Repost to share with other course creators and learning designers in your network.

🤿 𝗜 𝘁𝗼𝗼𝗸 𝗮 𝗱𝗲𝗲𝗽 𝗱𝗶𝘃𝗲 𝗶𝗻𝘁𝗼 𝗚𝗼𝗼𝗴𝗹𝗲 𝗡𝗼𝘁𝗲𝗯𝗼𝗼𝗸𝗟𝗠’𝘀 𝗔𝘂𝗱𝗶𝗼 𝗢𝘃𝗲𝗿𝘃𝗶𝗲𝘄 —𝗵𝗲𝗿𝗲 𝗮𝗿𝗲 𝗺𝘆 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀: In case you missed it, NotebookLM is Google’s Gemini AI-powered assistant that helps synthesize information from various sources. The new Audio Overview feature enhances this by turning your content into podcast-style discussions. 𝗪𝗵𝘆 𝗶𝘀 𝘁𝗵𝗶𝘀 𝗮 𝗯𝗶𝗴 𝘂𝗽𝗱𝗮𝘁𝗲? 👉🏻 Gemini 1.5 AI: This advanced model offers exceptional comprehension and easily synthesizes complex or multiple sources of information. 👉🏻 Cutting-edge synthetic voice tech: For the first time, we hear two AI voices interacting naturally, complete with pauses, filler words, and even laughter. 𝗛𝗼𝘄 𝗰𝗮𝗻 𝗔𝘂𝗱𝗶𝗼 𝗢𝘃𝗲𝗿𝘃𝗶𝗲𝘄 𝗯𝗲 𝘂𝘀𝗲𝗱 𝗮𝘀 𝗮 𝗹𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝘁𝗼𝗼𝗹? 🔷 Social learning: Simulates natural conversation and promotes deeper understanding through accessible dialogue—much like how podcasts enhance learning. 🔷 Personalized learning: Provides flexibility, allowing learners to choose between reading or listening based on their preferences or situation. 🔷 Multimodal reinforcement: Delivers the same content in both text and audio formats, reinforcing comprehension and retention. 🔷 Accessibility: Enhances inclusivity for learners with varying needs or disabilities by offering different modes of content delivery. 🔷 Conversation analysis: The AI-generated dialogue offers a unique opportunity to analyze spoken language, especially through its (often exaggerated) use of conversational elements. 𝗣𝗼𝘀𝘀𝗶𝗯𝗹𝗲 𝘂𝘀𝗲𝘀 𝗳𝗼𝗿 𝘁𝗵𝗲 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗹𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗰𝗹𝗮𝘀𝘀𝗿𝗼𝗼𝗺: 1️⃣ 𝗙𝗼𝗰𝘂𝘀 𝗼𝗻 𝗹𝗶𝘀𝘁𝗲𝗻𝗶𝗻𝗴 𝘀𝗸𝗶𝗹𝗹𝘀: Upload an article, let NotebookLM generate an audio discussion, and use it for targeted listening practice. 2️⃣ 𝗣𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗲𝗱 𝗽𝗼𝗱𝗰𝗮𝘀𝘁 𝘀𝘂𝗺𝗺𝗮𝗿𝗶𝗲𝘀: Create podcast-style summaries of learners’ essays or research notes, then add a follow-up task for learners to elaborate on the audio. 3️⃣ 𝗦𝘁𝘂𝗱𝘆 𝗴𝘂𝗶𝗱𝗲 𝗳𝗼𝗿 𝗿𝗲𝘃𝗶𝘀𝗶𝗼𝗻: Upload class notes and generate an audio summary. Learners can listen while commuting or during other passive activities. 4️⃣ 𝗖𝘂𝗿𝗮𝘁𝗲𝗱 𝗿𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀: Create a custom podcast for learners by gathering various sources on a specific topic of interest. 5️⃣ 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝘃𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀: Learners research a topic using various sources, input the collected data, generate an audio overview, and collaborate to refine and present their findings to the class. What other innovative ways could you imagine using AI like NotebookLM in L&D and education? Let me know your thoughts or share your experiences with this tool! 𝗡𝗼𝘁𝗲: Let’s not forget the value of real-time human interaction. For instance, why not challenge learners to record an actual podcast as part of their learning process?

Stop reinventing the wheel. Start curating it. Did you know your people could be losing 10 hours a week chasing the information they actually need? That’s a full day of productivity evaporated.¹ Here’s the twist: the problem isn’t too little content—it’s too much. L&D teams are grinding to build everything from scratch, while marketing quietly perfected a counter—content curation.₂ Let learners swim, don’t drown. Instead of fueling the content fire, what if we became expert filters—trusted guides who deliver exactly what matters, exactly when it matters? The CURATED model lays it out: - Clarify goals - Unearth existing gems - Refine them for relevance - Arrange in bite-sized flows - Transform for the learner’s experience - Engage with interaction and feedback - Deliver—and loop back with iteration Marketers did this first. They flipped chaos into clarity by treating content like curated conversation, not clutter. Make curation your superpower: Your next move is simple: Flip your mindset: Forget content first—think learner first. Start your CURATED experiment: Even one curated list or learning calendar can demonstrate impact. Invite your SMEs to team up: Build a network of internal curators, not lone creators. In a world drowning in content chaos, be the beacon. Don’t add to the flood—be the one who shines the way. Your learners don’t need another content dump. They need a compass. https://lnkd.in/gCjUCY8h What’s one topic your team could stop building and start curating this month?

Google recently dropped 'Learn Your Way'. It’s aimed at schools, but is a sign of things to come for workplace learning. With 'Learn Your Way,' students pick a grade level, choose an interest (currently from a selection), and your personalised textbook is generated with examples, illustrations, quizzes, etc that reference your level & interests. 🔗 Try it here: https://lnkd.in/gTux6cvU 📄 Or dive into the research paper: https://lnkd.in/gqsvGkdj ‘Learn Your Way’ leverages student interests to maintain engagement, but that’s a short step from selecting industry, job level, or even a current project for ultra relevant examples and case studies. For me, this highlights the importance of curating 'trusted core content' that can be personalised and contextualised for each individual via AI systems, then scaffolding that with on the job experiences, and networks. The (over simplified) partnership might look like this: Knowledge Management (Human led) + Digital Learning Design (AI led) + Experience Design (Human led) = Accelerated Learning In other words, if your L&D career is based on content creation, it’s time to upskill into providing real-world experiences, connections, and on-the-job applications. What do you think? I’m keen to hear your reflections on 'Learn Your Way' and the trend it represents in the comments.

The Bertrand Education Group (B.E.G) believes education should be as adaptive as the students we serve. 🚀 "Every student learns differently, but our system treats them the same." This insight drove the development of PrepAI's personalized approach to education. Our Impact in the $187B EdTech Market: 1. Personalized Learning at Scale: - Tailored pathways for each learner - 23% improvement in academic performance - Adaptive content delivery - Real-time progress monitoring 2. Empowering Educators: - 37% initial efficiency gains for teachers - Reduced administrative workload - Enhanced instructional focus - Data-driven teaching strategies 3. AI-Powered Assessment Innovation: - Generate adaptive assessments 12x faster - Reduce manual workload by 79% - Comprehensive student insights - Continuous improvement metrics 4. Global Access & Economic Mobility: - Implementation across U.S. and Indian institutions - Cross-cultural adaptability - Democratized quality education - Pathways to opportunity regardless of background In partnership with Microsoft for Startups and Qatar Foundation, we're transforming education through technology that unlocks human potential. What educational challenges do you believe AI can help solve? Share your thoughts below. #EdTech #AI #Innovation #PrepAI #PersonalizedLearning #FutureOfEducation