Problems to prototypes

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: An AI Debate Coach for Evidence, Reasoning, and Voice

CAPABILITY GAP:
Students often need far more low-stakes rehearsal in argumentation and public speaking than a teacher or coach can provide. Generic generative AI can produce polished arguments for learners, however, which risks replacing the very reasoning and voice the practice is meant to develop.

FIRST PROTOTYPE:
Create an AI rehearsal partner that listens to or reads a learner's argument, identifies the claim-evidence-reasoning structure, poses a counterargument, and gives process-focused feedback. It should never write the final speech by default and should make uncertainty and source quality visible.

REPRESENTATIVE USE CASE:
A learner preparing a school debate gives a two-minute opening statement. The coach asks one skeptical follow-up, highlights an unsupported inference, and invites a revised response before showing any model language.

LEARNING / HUMAN-SYSTEM FRAME:
The core mechanism is deliberate, repeated practice with specific feedback and self-explanation. Human judgment remains central: a teacher or coach sets norms, evaluates rhetoric and ethics, and helps learners interpret feedback rather than treating an AI score as authoritative.

QUESTIONS ALREADY IDENTIFIED:
1. Does repeated AI rehearsal improve live performance with a human audience?
2. Which feedback prompts strengthen reasoning without homogenizing student voice?
3. How should the system handle controversial topics or biased source material?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to construct, defend, and revise an evidence-based argument in real time, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, coach or teacher, peers/audience, source materials, AI rehearsal partner, and debate norms. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: argument quality, response to counterarguments, source use, oral confidence, transfer to live debate, and reliance on AI wording. A key disconfirming signal is: fluency improves but original reasoning, source judgment, or authentic voice declines.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A District AI Use Framework That Protects Learning

CAPABILITY GAP:
Schools are adopting generative AI faster than many districts can define what responsible use looks like. Simple bans or blanket permission do not distinguish between AI that supports learning, AI that replaces learning, and AI that creates unacceptable privacy or equity risks.

FIRST PROTOTYPE:
Develop a policy-and-practice sandbox that classifies proposed AI uses by learner age, instructional purpose, data sensitivity, human oversight, and the capability at risk. It generates a draft classroom protocol and student-facing explanation, while routing high-risk uses for human review.

REPRESENTATIVE USE CASE:
A district considers AI feedback on middle-school writing. The sandbox distinguishes brainstorming, formative feedback, and full-text generation, then asks what evidence must remain visibly the student's own.

LEARNING / HUMAN-SYSTEM FRAME:
The design treats AI literacy and governance as part of the learning system, not as a separate compliance layer. Policy should preserve human agency, make expectations teachable, and align tool use with a specific learning purpose and observable evidence.

QUESTIONS ALREADY IDENTIFIED:
1. Which uses should be encouraged, constrained, or prohibited by age and task?
2. How will teachers and students learn the policy rather than merely acknowledge it?
3. What evidence would show that a rule improves learning instead of only reducing incidents?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to make and explain responsible decisions about AI use in learning, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: students, teachers, families, district leaders, approved tools, data governance, curriculum, and local policy. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: policy comprehension, consistency of decisions, learning-task integrity, privacy incidents, teacher workload, and equity of access. A key disconfirming signal is: rules are followed superficially but teachers and students cannot reason about new AI situations.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Family–School AI Connector for Literacy Support

CAPABILITY GAP:
Literacy support is distributed across classrooms, homes, community programs, and digital tools, yet families and educators often lack a shared picture of what the learner is practicing and how to help. AI personalization can add another disconnected layer if it is not designed around relationships and instructional goals.

FIRST PROTOTYPE:
Create a teacher-configured family learning companion that translates a weekly literacy target into short home activities, optional multilingual explanations, and questions families can use to notice progress. The system reports patterns back to the teacher without treating home behavior as a surveillance signal.

REPRESENTATIVE USE CASE:
A learner is practicing inference in short texts. The family companion suggests a five-minute conversation using a familiar story or local event, then lets the family record what was easy or confusing.

LEARNING / HUMAN-SYSTEM FRAME:
Learning support should connect school goals to meaningful contexts while respecting family knowledge and cultural assets. The AI is a communication and scaffolding layer; the teacher and family remain the interpreters of learner needs.

QUESTIONS ALREADY IDENTIFIED:
1. What information is genuinely useful to families?
2. How much data should flow back to school?
3. Does the prototype strengthen literacy performance and family participation without increasing burden or inequity?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to apply a literacy strategy across school and home contexts, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, family, teacher, community context, curriculum target, AI companion, and privacy boundaries. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: strategy use, reading performance, family participation, teacher usefulness, burden, and subgroup access. A key disconfirming signal is: communication volume increases without improvement in learner strategy use or family trust.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: Low-Bandwidth AI Support for Small and Rural Schools

CAPABILITY GAP:
AI-enabled learning tools are often designed around reliable broadband, abundant devices, and specialist support. Those assumptions can make the technology least usable in the schools where access to expert instruction and staff capacity are already constrained.

FIRST PROTOTYPE:
Prototype a low-bandwidth, teacher-first AI assistant that works from locally cached curriculum materials, generates offline-ready practice and explanations, and synchronizes only when connectivity is available. High-stakes decisions stay with educators, and the design includes a non-AI fallback.

REPRESENTATIVE USE CASE:
A small secondary school with intermittent internet uses the assistant to prepare differentiated science practice packets and teacher discussion prompts from an approved local curriculum set.

LEARNING / HUMAN-SYSTEM FRAME:
The goal is not 'AI access' but improved instructional capacity under real constraints. Human-centered design requires explicit modeling of infrastructure, teacher time, local curriculum, and data risk alongside learner needs.

QUESTIONS ALREADY IDENTIFIED:
1. What useful functions still work with weak connectivity?
2. Does the tool reduce teacher preparation time without lowering instructional quality?
3. Are learning gains comparable when the AI is intermittently available?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to deliver and adapt high-quality instruction despite limited infrastructure, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learners, teachers, local curriculum, devices, connectivity, school leadership, AI assistant, and offline workflow. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: teacher time, task quality, learner performance, uptime dependence, equity of access, and fallback success. A key disconfirming signal is: the system saves time only when connectivity and technical support are already strong.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: Mastery Pathways for Mixed-Readiness Classrooms

CAPABILITY GAP:
A single classroom can contain wide variation in prior knowledge and pace, while teachers have limited time to diagnose every learner and create individualized practice. AI personalization can help, but only if it is tied to clear mastery evidence rather than opaque engagement scores.

FIRST PROTOTYPE:
Build a teacher-facing mastery pathway tool that links each learning objective to diagnostic tasks, prerequisite skills, practice options, and observable mastery evidence. AI can propose next activities, but teachers see and can override the reasoning.

REPRESENTATIVE USE CASE:
In a mathematics unit, one learner receives prerequisite fraction practice, another receives a transfer problem, and both must demonstrate the same target concept through an explanation and novel problem.

LEARNING / HUMAN-SYSTEM FRAME:
Personalization should adapt support while holding the capability target and evidence standard stable. Feedback, metacognitive monitoring, and opportunities to transfer learning matter more than merely accelerating task completion.

QUESTIONS ALREADY IDENTIFIED:
1. Which evidence is sufficient to advance a learner?
2. How transparent must the recommendation logic be for teachers and families?
3. Does the system narrow or widen opportunity gaps over time?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to reach and demonstrate mastery from different starting points, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learners, teacher, curriculum progression, diagnostic tasks, AI recommender, and classroom schedule. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: mastery, time-to-mastery, transfer, teacher override patterns, opportunity gaps, and learner self-monitoring. A key disconfirming signal is: the system optimizes completion pace without improving durable mastery or equity.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Practice Lab for Teacher AI Adoption

CAPABILITY GAP:
AI adoption in schools often depends on a few enthusiastic users while other educators receive tool demos without enough time to build judgment about when AI is useful, risky, or instructionally counterproductive.

FIRST PROTOTYPE:
Build a scenario-based professional learning lab where teachers practice making AI-use decisions, try bounded tools on authentic planning tasks, see failure cases, and receive feedback tied to an explicit competency framework and local policy.

REPRESENTATIVE USE CASE:
A teacher chooses whether to use AI to generate feedback on student writing, then compares three configurations: unrestricted generation, rubric-grounded suggestions, and teacher-only planning support.

LEARNING / HUMAN-SYSTEM FRAME:
Teacher learning should involve application, reflection, feedback, and opportunities to revise—not one-time compliance training. The goal is professional judgment and human agency in a teacher–AI–student system.

QUESTIONS ALREADY IDENTIFIED:
1. Which competencies predict responsible classroom use?
2. What kinds of practice change teacher behavior rather than only attitudes?
3. How should local policy and subject-area pedagogy alter the scenarios?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to make context-sensitive instructional decisions about AI use, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: teacher, students, curriculum, school policy, AI tools, professional learning team, and leadership. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: scenario judgment, classroom implementation quality, policy alignment, teacher confidence calibration, and student learning evidence. A key disconfirming signal is: teachers report higher confidence but make no better instructional or risk-management decisions.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: XR + AI for Safe Practice of Complex Workflows

CAPABILITY GAP:
Complex technical workflows are hard to learn from manuals and slide decks because learners must coordinate spatial, procedural, and decision knowledge. High-fidelity simulation can be expensive, while low-cost digital training often lacks realistic practice.

FIRST PROTOTYPE:
Prototype a lightweight 3D or XR scenario with an AI coach that observes task sequence, asks the learner to predict the next step, and provides graduated feedback. The first version should simulate only one high-value workflow rather than build a broad virtual world.

REPRESENTATIVE USE CASE:
A new employee practices configuring a piece of enterprise equipment in a virtual environment, including one realistic fault that requires diagnosis rather than rote sequence-following.

LEARNING / HUMAN-SYSTEM FRAME:
Multimedia should reduce extraneous processing and make invisible relationships visible, while simulation enables repeated practice without real-world risk. Feedback should move from guidance toward independence as competence grows.

QUESTIONS ALREADY IDENTIFIED:
1. Which elements require spatial simulation and which can remain 2D?
2. Does immersive practice transfer to the real workflow?
3. Is added realism worth the cost and cognitive load?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to execute and troubleshoot a complex workflow safely and independently, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, trainer, workflow, equipment model, XR interface, AI coach, and workplace safety constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: sequence accuracy, diagnosis quality, transfer to real task, error recovery, time-to-competence, and simulator cost. A key disconfirming signal is: the simulation feels realistic but does not improve real-world performance beyond simpler practice.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Clinical Reasoning + AI Calibration Coach

CAPABILITY GAP:
Clinical trainees increasingly encounter AI-supported decisions, but learning to use AI safely requires more than knowing what a tool can output. Learners must integrate foundational knowledge, uncertainty, patient context, and the possibility that the AI is wrong.

FIRST PROTOTYPE:
Create a clinical reasoning simulator in which an AI assistant sometimes offers useful, incomplete, or misleading suggestions. The learner must state an independent assessment, identify what evidence would change the decision, and explicitly answer 'what might I be missing?' before seeing expert feedback.

REPRESENTATIVE USE CASE:
A simulated patient case includes an AI-generated differential diagnosis that omits a less common but consequential possibility. The learner decides whether to accept, challenge, or investigate the suggestion.

LEARNING / HUMAN-SYSTEM FRAME:
Simulation with deliberate practice can support complex professional judgment when feedback is explicit and repeated. Metacognitive prompts make uncertainty and self-monitoring visible; the AI is deliberately treated as fallible evidence, not an authority.

QUESTIONS ALREADY IDENTIFIED:
1. How often should the AI be wrong to teach calibration without creating distrust?
2. Which measures capture reasoning quality rather than answer matching?
3. Does practice change behavior in authentic clinical settings?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to integrate AI advice into independent, evidence-based professional judgment, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: trainee, supervisor, patient case, clinical knowledge, AI assistant, institutional policy, and safety constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: diagnostic reasoning, appropriate challenge of AI, uncertainty calibration, transfer, safety-critical misses, and explanation quality. A key disconfirming signal is: learners learn to distrust or obey the AI categorically instead of calibrating trust to evidence.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: An AI Academic Navigator for Two-Year Colleges

CAPABILITY GAP:
Students in open-access and two-year institutions often navigate fragmented academic, financial, technology, and support systems while balancing work and family responsibilities. Generic AI can provide 24/7 help but may confidently invent institutional rules or conceal where human support is necessary.

FIRST PROTOTYPE:
Create a retrieval-grounded academic navigator using only approved institutional sources. It explains options, checks understanding, and escalates financial aid, disability, mental health, or policy-sensitive questions to people rather than making decisions.

REPRESENTATIVE USE CASE:
A working adult asks how dropping one course might affect progress. The navigator identifies the relevant institutional resources, explains which variables matter, and routes the student to the appropriate office for a binding answer.

LEARNING / HUMAN-SYSTEM FRAME:
The learning goal is institutional self-navigation: knowing what to ask, how to interpret information, and when to seek human help. Human-centered AI design requires calibrated confidence and clear boundaries.

QUESTIONS ALREADY IDENTIFIED:
1. Which questions are safe to automate?
2. Does the navigator improve successful follow-through, not merely chat satisfaction?
3. Are nontraditional learners using and trusting the system equitably?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to navigate institutional systems and make informed next-step decisions, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: student, advisors, institutional services, approved knowledge base, AI navigator, and escalation workflow. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: correct action, successful referral, resolution time, hallucination rate, trust calibration, and access by learner group. A key disconfirming signal is: usage is high but students still miss deadlines, choose wrong actions, or avoid human support when needed.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: Beyond Translation: A Multilingual Family Communication Studio

CAPABILITY GAP:
Translating an English resource word-for-word can leave examples, tone, assumptions, and calls to action culturally distant from multilingual learners and families. Educators need a faster way to adapt communication without treating language as the only barrier.

FIRST PROTOTYPE:
Create an educator-facing adaptation studio that produces bilingual drafts, identifies culture- or context-dependent references, proposes locally relevant examples, and asks the educator to confirm meaning before publication. Families can optionally flag confusing or unnatural wording.

REPRESENTATIVE USE CASE:
A school family guide about mathematics practice is adapted into two languages, but also replaces unfamiliar examples, clarifies school-specific terms, and offers an audio version for families who prefer listening.

LEARNING / HUMAN-SYSTEM FRAME:
Effective multilingual learning builds on home language and cultural assets rather than using translation only as remediation. Multiple representations improve access, while human review protects meaning and relationship.

QUESTIONS ALREADY IDENTIFIED:
1. What makes an adaptation feel authentic rather than merely fluent?
2. How should family feedback alter future drafts?
3. Does better communication change participation and learner outcomes?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to understand and act on learning-related communication across languages and cultural contexts, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, family, educator, home language, school context, communication artifact, and AI adaptation studio. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: comprehension, follow-through, family trust, revision rate, educator time, and cross-language consistency. A key disconfirming signal is: translation quality improves but family understanding, participation, or trust does not.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Curiosity-Preserving AI Mentor

CAPABILITY GAP:
Personalized AI mentors may optimize completion, hints, and immediate success while unintentionally reducing exploration, learner agency, or tolerance for difficulty. A system that predicts the learner too aggressively can become a path optimizer instead of a partner in discovery.

FIRST PROTOTYPE:
Design an AI mentor that begins with learner-chosen goals, uses short challenge cycles, offers optional creative pathways, and periodically asks the learner to decide what to pursue next. Human mentors can see the rationale for interventions and override the system.

REPRESENTATIVE USE CASE:
A teenager learning game design alternates between coding challenges and visual design choices. The AI notices repeated debugging difficulty but offers two next-step paths rather than automatically simplifying the task.

LEARNING / HUMAN-SYSTEM FRAME:
The design supports autonomy, competence, metacognition, and human oversight. Personalization is treated as a hypothesis about support, not a hidden judgment about ability or motivation.

QUESTIONS ALREADY IDENTIFIED:
1. How do we measure curiosity without reducing it to clicks or facial signals?
2. When should the mentor challenge rather than accommodate?
3. Does learner choice improve persistence and transfer over time?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to pursue increasingly complex learning with agency, reflection, and sustained curiosity, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, human mentor, chosen project, AI mentor, peer/community context, and progression model. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: independent project complexity, persistence, self-explanation, choice patterns, transfer, and mentor override. A key disconfirming signal is: the mentor keeps learners active but narrows exploration or makes challenge avoidance easier.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Context-Rich AI Language Practice Companion

CAPABILITY GAP:
Language apps can make practice frequent and convenient, yet generic drills may not prepare immigrant and multilingual learners for the specific academic, social, and institutional situations where language must be used.

FIRST PROTOTYPE:
Create a teacher-configurable AI language companion grounded in real school and community tasks. It supports speech, text, images, home-language clarification, and repeated practice while keeping the target communicative function explicit.

REPRESENTATIVE USE CASE:
A newcomer practices asking a teacher for clarification, interpreting an assignment instruction, and contributing one idea to a group discussion before trying the same tasks in class.

LEARNING / HUMAN-SYSTEM FRAME:
The design connects language learning to authentic use, builds on the learner's existing linguistic resources, and offers multiple representations and response modes. AI provides practice volume; educators determine goals and interpret progress.

QUESTIONS ALREADY IDENTIFIED:
1. Which simulated situations transfer most strongly to real participation?
2. When is home-language support helpful versus overused?
3. Does practice increase confidence and actual classroom communication?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to use a second language successfully in authentic academic and social situations, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, teacher, peers, home language, classroom/community tasks, AI language companion, and school norms. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: communicative success, vocabulary/structure growth, classroom participation, transfer, confidence calibration, and language-support use. A key disconfirming signal is: app performance improves without more successful communication in real settings.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: From Engagement Data to Better STEM Learning

CAPABILITY GAP:
Many educational technologies can tell us that a learner clicked, watched, or stayed on task, but not whether the learner developed a transferable capability. In project-based STEM settings, engagement data can easily become a proxy for learning rather than evidence of it.

FIRST PROTOTYPE:
Build an AI-assisted lesson design partner that starts with a capability target, generates short project challenges, and recommends evidence-producing learner actions. The AI proposes adaptations, but the educator chooses the intervention and the evidence standard.

REPRESENTATIVE USE CASE:
A middle-school robotics unit asks learners to diagnose a robot that fails only under certain sensor conditions. The system offers progressively different representations or hints only after the learner explains a hypothesis.

LEARNING / HUMAN-SYSTEM FRAME:
The prototype should privilege constructive and interactive activity over passive consumption, use feedback tied to the task and reasoning process, and preserve productive struggle. The human–learner system includes the learner, teacher, peers, task, interface, classroom constraints, and the evidence used to decide whether capability actually improved.

QUESTIONS ALREADY IDENTIFIED:
1. What learner action would count as evidence of debugging capability rather than engagement?
2. When should the AI withhold help?
3. Do adaptations improve transfer to a new robotics problem, and for whom?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to diagnose and explain failures in a novel STEM task, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, teacher, peers, project task, AI design partner, classroom time, and available devices. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: independent diagnosis accuracy, quality of explanation, transfer to a novel task, time-to-help, and subgroup differences. A key disconfirming signal is: engagement rises while independent reasoning or transfer stays flat.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: AI Role-Play for Evidence-Based Global Inquiry

CAPABILITY GAP:
Open-web research can reward shallow answer-finding, especially when learners are new to a topic or working in a second language. AI can accelerate that pattern unless it is deliberately designed to make learners compare evidence, adopt perspectives, and justify decisions.

FIRST PROTOTYPE:
Build a teacher-configured AI role-play environment in which learners interact with simulated stakeholders who have different goals and incomplete information. The system requires evidence citations and prompts learners to revise claims when a source is weak or a stakeholder perspective changes.

REPRESENTATIVE USE CASE:
In a global energy unit, teams advise a fictional regional summit. An AI investor, community representative, and environmental scientist each challenge the team's proposal from a different perspective.

LEARNING / HUMAN-SYSTEM FRAME:
Role-play can move learners from passive information retrieval toward constructive and interactive sense-making. Multilingual supports should scaffold access without lowering the reasoning demand, and the teacher should control the task, sources, and acceptable use of AI.

QUESTIONS ALREADY IDENTIFIED:
1. Does role-play produce better source evaluation than ordinary web research?
2. Which multilingual supports increase participation without masking misconceptions?
3. Can learners transfer perspective-taking to a new issue?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to synthesize evidence and adapt an argument to multiple stakeholder perspectives, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: multilingual learners, teacher, peers, curated sources, AI stakeholders, language supports, and classroom norms. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: source quality, evidence integration, perspective-taking, revision quality, participation equity, and transfer. A key disconfirming signal is: learners produce more elaborate answers but rely on AI-generated reasoning or weak sources.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: Scenario-Based Learning for Complex Professional Knowledge

CAPABILITY GAP:
Professionals can memorize rules, policies, or procedures yet struggle when real work presents ambiguity, tradeoffs, or incomplete information. Static training often teaches what a rule says but not when and how to apply it.

FIRST PROTOTYPE:
Create an AI scenario studio that converts a bounded body of expert-approved content into branching cases. Learners make a decision, explain why, see consequences, and receive feedback tied to the relevant principle rather than a generic correctness score.

REPRESENTATIVE USE CASE:
A new analyst must choose how to handle a client request that conflicts with an internal policy. The case changes after the learner's first decision, forcing a second judgment under new constraints.

LEARNING / HUMAN-SYSTEM FRAME:
The design targets transfer by varying context while holding the underlying principle stable. Multiple representations and worked examples can reduce unnecessary cognitive load early, then fade as learners gain competence.

QUESTIONS ALREADY IDENTIFIED:
1. Which parts of professional judgment can be safely simulated?
2. What is the right balance between worked examples and independent cases?
3. Does performance transfer to unfamiliar cases rather than only improve on practiced patterns?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to apply a complex rule or principle to an unfamiliar professional situation, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, subject-matter expert, supervisor, policy corpus, AI scenario engine, and workplace constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: decision quality, explanation quality, error type, transfer to novel cases, time-to-competence, and expert agreement. A key disconfirming signal is: case scores improve but learners cannot explain or apply the principle outside the simulator.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: AI With Friction: Helping Learners Think Before They Ask

CAPABILITY GAP:
Generative AI can make schoolwork easier while also removing the retrieval, explanation, and revision that make learning durable. The design problem is not simply whether students use AI, but whether AI appears before or after the learner has done cognitively useful work.

FIRST PROTOTYPE:
Build a 'productive friction' assistant that asks for an initial attempt, confidence rating, or explanation before offering help. It provides graduated hints, asks the learner to critique its answer, and periodically withholds generation in favor of retrieval or self-explanation.

REPRESENTATIVE USE CASE:
A university learner asks for help with a difficult reading. Instead of summarizing immediately, the assistant asks for three claims the learner noticed, then helps compare those claims to the text.

LEARNING / HUMAN-SYSTEM FRAME:
The intervention treats metacognition, retrieval, and constructive engagement as capabilities worth protecting. AI should support monitoring and feedback while leaving the learner responsible for sense-making.

QUESTIONS ALREADY IDENTIFIED:
1. How much friction is enough to improve learning without driving abandonment?
2. Which tasks benefit from delayed AI help?
3. Can the system detect overreliance without inferring motivation or ability from weak behavioral proxies?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to monitor one's understanding and use AI without outsourcing core reasoning, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, instructor, assignment, source material, AI assistant, academic norms, and help-seeking choices. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: retention, explanation quality, independent task performance, help-seeking pattern, confidence calibration, and abandonment. A key disconfirming signal is: the design changes interaction patterns but not retention or independent performance.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: A Family Learning Studio for Early Learners

CAPABILITY GAP:
Young learners often experience school and home as separate learning environments even when family participation strongly shapes routines, language, and motivation. Families may receive generic activities that do not fit their language, culture, available materials, or the child's current learning goal.

FIRST PROTOTYPE:
Create an educator-controlled family activity studio that turns a weekly learning objective into short parent-child activities using common household materials, optional multilingual supports, and multiple ways for the child to respond.

REPRESENTATIVE USE CASE:
For an early mathematics goal, the studio generates a sorting and comparison game using kitchen objects, with a visual version, a story version, and a movement-based version.

LEARNING / HUMAN-SYSTEM FRAME:
The design draws on learner context, prior knowledge, multiple means of engagement and expression, and family cultural assets. AI generates options; educators and families choose what is meaningful and feasible.

QUESTIONS ALREADY IDENTIFIED:
1. Which adaptations actually increase participation rather than just choice?
2. How can family feedback inform instruction without becoming homework surveillance?
3. Do gains transfer back to classroom tasks?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to use an emerging academic concept across school and home activities, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: young learner, family, teacher, home materials, language context, curriculum goal, and AI activity studio. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: participation, concept performance, transfer to classroom, family usability, accessibility, and burden. A key disconfirming signal is: families receive more activities but participation or classroom transfer does not improve.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: A Comprehension-First Navigator for International Students

CAPABILITY GAP:
International students must act on dense, high-stakes institutional information that may be linguistically unfamiliar and distributed across many offices. A chatbot that merely retrieves policy text can increase speed without ensuring comprehension or appropriate escalation.

FIRST PROTOTYPE:
Build a retrieval-grounded navigator that explains approved institutional information in plain language, checks understanding, identifies when a question exceeds its authority, and routes the learner to a human office with the relevant context.

REPRESENTATIVE USE CASE:
A student asks what steps are required before an international internship. The navigator gives an institution-approved checklist, asks the student to restate the next action, and flags questions that require an advisor.

LEARNING / HUMAN-SYSTEM FRAME:
The key capability is not information access but correct interpretation and action. The system should reduce extraneous language burden while preserving important distinctions, uncertainty, and human accountability.

QUESTIONS ALREADY IDENTIFIED:
1. Does the tool improve correct action, not just answer satisfaction?
2. When should it stop and escalate?
3. Which simplifications help multilingual learners without distorting policy?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to interpret complex institutional information and take the correct next action, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: student, advisors, institutional policy corpus, AI navigator, language supports, and escalation channels. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: comprehension, correct next action, escalation accuracy, resolution time, trust calibration, and subgroup differences. A key disconfirming signal is: users report convenience but make the same or more procedural errors.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: An Inquiry Studio for Exam-Centric Learning

CAPABILITY GAP:
Systems optimized around exams can reward reproduction of correct answers while underdeveloping curiosity, experimentation, and the ability to formulate a problem. Adding AI can make answer production even easier unless the learning experience explicitly values inquiry.

FIRST PROTOTYPE:
Create an AI inquiry studio that converts required syllabus topics into locally meaningful investigations. Learners choose a question, predict an outcome, gather evidence, and defend a conclusion; the AI helps generate alternatives and critique reasoning but does not provide a final answer first.

REPRESENTATIVE USE CASE:
A chemistry topic becomes a neighborhood water-quality investigation in which learners select variables, interpret imperfect data, and explain what they still cannot conclude.

LEARNING / HUMAN-SYSTEM FRAME:
The design emphasizes active and constructive learning, learner agency, and metacognitive reflection. Required content remains visible, but success includes the quality of questions, evidence use, and transfer—not only answer accuracy.

QUESTIONS ALREADY IDENTIFIED:
1. Can inquiry coexist with high-stakes exam requirements?
2. Which kinds of learner choice support motivation without overwhelming novices?
3. Does the approach improve transfer and curiosity beyond the project itself?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to formulate, investigate, and revise a question using disciplinary evidence, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, teacher, syllabus, local context, peers, AI inquiry partner, and assessment constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: question quality, evidence use, transfer, content mastery, persistence, and learner agency. A key disconfirming signal is: projects feel engaging but required knowledge or independent inquiry does not improve.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: An EdTech Alignment Auditor: Is the App Actually Teaching?

CAPABILITY GAP:
Students can be highly fluent with devices and learning apps while the activity itself remains weakly connected to the intended lesson. Schools often evaluate tools by engagement, usability, or vendor claims rather than by the learner actions and evidence the tool produces.

FIRST PROTOTYPE:
Create an AI-assisted alignment auditor that takes a learning objective, lesson plan, and description or capture of a digital activity, then maps the activity to the cognitive work learners are actually asked to do. It suggests a stronger task or teacher prompt when alignment is weak.

REPRESENTATIVE USE CASE:
A literacy app asks learners to tap vocabulary definitions. The auditor identifies that the lesson target is inferential reading and proposes a short evidence-based explanation task before the app is used.

LEARNING / HUMAN-SYSTEM FRAME:
The design makes the learner action—not the technology—the unit of analysis. It distinguishes passive, active, constructive, and interactive engagement and asks whether feedback is connected to the learning goal.

QUESTIONS ALREADY IDENTIFIED:
1. Can the auditor reliably distinguish engaging activity from useful cognitive work?
2. How should teacher judgment override automated mappings?
3. Does using the auditor improve lesson-level evidence of learning?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to select and configure technology that elicits the intended learner thinking, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: teacher, learners, lesson objective, digital tool, AI auditor, curriculum, and classroom workflow. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: objective-task alignment, quality of learner evidence, teacher decision quality, student performance, and tool abandonment. A key disconfirming signal is: teachers produce better-looking lesson plans but learner actions and outcomes remain unchanged.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: An Accessibility-First Learning Adaptation Assistant

CAPABILITY GAP:
Educators supporting learners with varied academic, communication, sensory, and behavioral needs often spend substantial time adapting the same material into accessible forms. Generic AI can speed adaptation but may also make unsupported assumptions about disability or lower the intellectual demand.

FIRST PROTOTYPE:
Create an educator-controlled adaptation assistant that produces multiple representations, response modes, scaffolds, and accessibility checks from a common learning target. It asks what barrier is being removed and preserves the intended cognitive demand.

REPRESENTATIVE USE CASE:
A social studies task can be rendered with simplified navigation, visual supports, text-to-speech-ready structure, and alternative response formats while keeping the same evidence-based reasoning goal.

LEARNING / HUMAN-SYSTEM FRAME:
Universal Design for Learning suggests designing options for access, engagement, and expression without treating a learner label as a prescription. The human system includes special educators, general educators, families, support staff, tools, and the learner's own preferences.

QUESTIONS ALREADY IDENTIFIED:
1. Which adaptations remove access barriers versus reduce rigor?
2. How should learner preference be captured?
3. Can the tool decrease educator workload while improving participation and performance?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to demonstrate the same target capability through accessible pathways, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, special and general educators, family/support team, curriculum, assistive technology, and AI adaptation assistant. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: task access, target performance, participation, learner preference, educator time, and inappropriate simplification rate. A key disconfirming signal is: materials become easier to access but the target capability is diluted or mismeasured.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: Closing the Knowing–Doing Gap in Professional Services

CAPABILITY GAP:
Early-career professionals can know the rules and still struggle to use them under time pressure, ambiguity, and interaction with clients or senior colleagues. Remote work can further reduce opportunities to observe expert judgment in action.

FIRST PROTOTYPE:
Create an AI-supported case simulator for a regulated professional-services context. Learners make a judgment, draft or speak a response, receive a simulated stakeholder reaction, and then compare their reasoning with expert-annotated alternatives.

REPRESENTATIVE USE CASE:
A junior professional receives a time-sensitive client request with incomplete information. The simulator forces a decision about what to ask, what to escalate, and what can safely be done now.

LEARNING / HUMAN-SYSTEM FRAME:
The prototype is built around transfer, deliberate practice, feedback, and reflection on expert reasoning. It should expose decision structure without pretending there is always one correct answer.

QUESTIONS ALREADY IDENTIFIED:
1. Which decisions are representative enough to practice safely?
2. How much expert annotation is required to keep AI feedback trustworthy?
3. Does simulator performance predict work-product quality on new matters?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to apply professional knowledge to ambiguous, time-constrained situations, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, supervisor, client/stakeholder, professional standards, AI simulator, and real workflow constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: decision quality, escalation judgment, rationale, transfer to new cases, supervisor rework, and confidence calibration. A key disconfirming signal is: learners master simulator patterns without improving judgment or reducing supervisor rework.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: Turning Tacit Expertise Into Performance Support

CAPABILITY GAP:
Organizations frequently rely on experienced staff to create ad hoc slide-based training, leaving valuable tacit knowledge fragmented and difficult to transfer. New employees may receive information without enough guided practice in the decisions that matter on the job.

FIRST PROTOTYPE:
Build an AI-assisted performance-support builder that interviews subject-matter experts, extracts recurring decisions and failure modes, and turns them into short scenarios, checklists, and job aids for expert review.

REPRESENTATIVE USE CASE:
An operations team captures how experienced staff triage an ambiguous incoming request, then turns that logic into a five-minute scenario and a one-page decision aid for new hires.

LEARNING / HUMAN-SYSTEM FRAME:
The design separates information that belongs in a job aid from judgment that requires practice. Learning is evaluated in the workflow, not by slide completion, and experts retain authority over the extracted model.

QUESTIONS ALREADY IDENTIFIED:
1. What knowledge can be externalized reliably from expert interviews?
2. Which decisions need practice rather than documentation?
3. Does the prototype reduce time-to-readiness and avoid propagating expert habits that are outdated or context-specific?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to perform recurring workplace decisions using expert-informed strategies, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: novice, expert, manager, workflow, existing documents, AI knowledge-capture tool, and performance-support artifacts. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: time-to-readiness, decision quality, expert review corrections, job-aid use, transfer, and rework. A key disconfirming signal is: content production accelerates but the extracted guidance does not improve real work.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: A Multimodal Concept Reframer

CAPABILITY GAP:
When a learner is stuck, repeating the same explanation with more words often fails because the representation—not the amount of explanation—is the barrier. Teachers and tutors cannot always generate several high-quality alternative representations in the moment.

FIRST PROTOTYPE:
Create an AI concept reframer that can present the same idea through a diagram, concrete example, analogy, verbal explanation, or guided question sequence. It then asks the learner to explain the concept back and uses that response to select the next representation.

REPRESENTATIVE USE CASE:
A learner struggling with a biological process receives a simple causal diagram and then a concrete analogy; the system asks the learner to map each part of the analogy back to the actual mechanism.

LEARNING / HUMAN-SYSTEM FRAME:
Multimedia and multiple representations can support meaning-making when they are coherent and tied to the same concept. The system should use learner explanations as evidence rather than infer a preferred 'learning style.'

QUESTIONS ALREADY IDENTIFIED:
1. Which representation changes understanding rather than just preference?
2. How can the system detect misconceptions in learner explanations?
3. Does the learner succeed later without the alternate representation?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to form a coherent mental model and explain a concept in one's own words, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, tutor/teacher, concept, representations, AI reframer, and follow-up task. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: explanation quality, misconception reduction, transfer, hint dependence, and representation sequence. A key disconfirming signal is: learners report clearer explanations but cannot solve or explain a novel application.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: An Inclusive Dialogue Studio for Difficult Conversations

CAPABILITY GAP:
Discussion of identity, culture, ethics, and contested social issues can produce rich learning, but participation is often uneven and instructors must manage language, power, emotion, and evidence simultaneously. An AI moderator could easily flatten disagreement or encode dominant cultural assumptions.

FIRST PROTOTYPE:
Build a pre-discussion and reflection studio that helps learners rehearse questions, examine language choices, identify evidence and assumptions, and consider missing perspectives. It should support the human facilitator rather than moderate or adjudicate the live conversation autonomously.

REPRESENTATIVE USE CASE:
Before a language-and-culture seminar, learners test how a phrase may be interpreted by different audiences, identify what evidence supports a claim, and prepare one genuine question for peers.

LEARNING / HUMAN-SYSTEM FRAME:
Culturally sustaining pedagogy treats learners' languages and identities as assets, while UDL emphasizes multiple ways to participate. The human facilitator remains responsible for norms, repair, and the ethical work of dialogue.

QUESTIONS ALREADY IDENTIFIED:
1. Which supports increase participation without scripting it?
2. How do we test for cultural bias in generated perspectives?
3. Does rehearsal improve listening, evidence use, and willingness to revise?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to engage respectfully and critically across differences using evidence and reflective communication, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learners, facilitator, peers, cultural/linguistic context, discussion norms, sources, and AI rehearsal studio. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: participation distribution, evidence use, listening/revision behaviors, perceived belonging, bias incidents, and facilitator workload. A key disconfirming signal is: discussion becomes more polished but less authentic, diverse, or willing to surface disagreement.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: An Access-First Digital Learning Experience Checker

CAPABILITY GAP:
Digital learning can expand access while still embedding barriers through navigation, language, assumptions about devices, inaccessible media, or rigid ways of responding. These barriers are often discovered after a course has already been built.

FIRST PROTOTYPE:
Create an AI-assisted preflight checker that reviews a lesson or course for likely access barriers, maps each barrier to the intended learner action, and proposes redesign options with an educator-facing rationale.

REPRESENTATIVE USE CASE:
A module requires a long video, timed discussion post, and complex drag-and-drop task. The checker identifies bandwidth, captioning, timing, and input-method barriers and suggests equivalent evidence-producing alternatives.

LEARNING / HUMAN-SYSTEM FRAME:
Accessibility is treated as part of learning design rather than remediation after failure. The system preserves the capability target while varying access, representation, and expression.

QUESTIONS ALREADY IDENTIFIED:
1. Which barriers can be detected reliably before learners encounter them?
2. How should lived learner feedback update the checker?
3. Does using it improve completion and learning without simplifying the intellectual work?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to access and demonstrate a learning target despite variation in learner context and technology, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, designer, instructor, content, platform, device/network context, accessibility services, and AI checker. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: barrier rate, completion, target performance, accommodation requests, designer time, and false-positive rate. A key disconfirming signal is: technical accessibility improves while meaningful participation or learning outcomes remain unchanged.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: A Transition Pathway Builder for Learners With Disabilities

CAPABILITY GAP:
Work-based and postsecondary transition opportunities often depend on staff knowledge, local partnerships, transportation, accommodations, and limited program capacity. As a result, access can be uneven even when learner goals are clearly documented.

FIRST PROTOTYPE:
Build a counselor-facing pathway builder that maps a learner's stated transition goals and demonstrated capabilities to a vetted set of local experiences, prerequisites, accommodations, and next-step skills. Recommendations remain explainable and require human approval.

REPRESENTATIVE USE CASE:
A student interested in audio production sees several pathways: a short school-based project, a local studio visit, and a supported internship, each with explicit preparation tasks and access requirements.

LEARNING / HUMAN-SYSTEM FRAME:
The system treats learner agency and real-world participation as central, while UDL and human review prevent a disability label from narrowing options. The capability model focuses on what the learner can demonstrate and what support the environment requires.

QUESTIONS ALREADY IDENTIFIED:
1. Are opportunity recommendations equitable across neighborhoods and learner profiles?
2. What data should never be used to rank learners?
3. Does the tool increase completed transition experiences and later independence?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to navigate and complete a personally meaningful transition experience, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, family, counselor, school, community partners, accommodations, transportation, and AI pathway builder. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: opportunity access, completion, learner choice, accommodation fit, equity across groups, and post-transition outcomes. A key disconfirming signal is: recommendations reproduce existing opportunity gaps or steer learners toward narrower options.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LXD
TITLE: A Curiosity Rhythm Coach, Not an Attention Monitor

CAPABILITY GAP:
Learners can sustain deep attention for personally meaningful activities yet disengage from school tasks. Systems often respond by measuring attention or adding rewards, which can mistake visible behavior for motivation and create culturally or neurologically biased interpretations.

FIRST PROTOTYPE:
Build an AI learning-design coach that helps educators create short challenge cycles with meaningful choice, visible progress, and varied media. It uses learner-selected goals and task performance—not eye tracking or body-language inference—to decide what to adjust.

REPRESENTATIVE USE CASE:
A language lesson is reorganized into a twelve-minute challenge, a choice of creative response, a quick feedback checkpoint, and a second challenge that increases difficulty if the learner chooses to continue.

LEARNING / HUMAN-SYSTEM FRAME:
Motivation is supported through autonomy, competence, relevance, and constructive activity rather than surveillance. Gamified elements are used only when they serve the learning process and should be tested for both motivational and cognitive effects.

QUESTIONS ALREADY IDENTIFIED:
1. Which design changes improve persistence and learning at the same time?
2. How should challenge length vary across learners without turning behavior into a label?
3. Do incentives support or displace curiosity?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to sustain purposeful effort and re-engage with challenging learning, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, teacher/designer, task, media, peer context, AI design coach, and classroom norms. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: persistence, target performance, voluntary re-engagement, autonomy/competence measures, and subgroup patterns. A key disconfirming signal is: time-on-task increases while learning, agency, or intrinsic interest declines.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS
TITLE: A Capability Map for Workforce Learning Systems

CAPABILITY GAP:
Large organizations often accumulate courses, credentials, and learning platforms without a shared model of the capabilities employees need to perform. Completion data then becomes easier to measure than readiness for real work.

FIRST PROTOTYPE:
Create a capability graph that links role outcomes to observable tasks, prerequisite knowledge, practice opportunities, and evidence. An AI coach can recommend practice from the graph and explain why, while managers and learning teams control the capability model.

REPRESENTATIVE USE CASE:
A new team lead is not assigned 'leadership content' broadly; the system identifies a need to run a difficult feedback conversation and recommends a short scenario, reflection, and live manager practice.

LEARNING / HUMAN-SYSTEM FRAME:
The learning architecture begins with performance and transfer rather than content inventory. AI supports navigation through the model, but human experts define what competent performance looks like and what evidence is credible.

QUESTIONS ALREADY IDENTIFIED:
1. Can the organization agree on observable capabilities across roles?
2. Which evidence predicts performance on the job?
3. Does recommendation improve readiness or simply increase learning activity?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to perform a defined workplace capability in context, not merely complete related training, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: employee, manager, learning team, role tasks, performance data, capability graph, and AI coach. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: task performance, transfer, time-to-readiness, manager agreement, recommendation uptake, and course reduction. A key disconfirming signal is: the graph becomes another taxonomy disconnected from actual work performance.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS
TITLE: Judgment-First Architecture for Online Programs

CAPABILITY GAP:
Online programs can scale procedural learning efficiently while leaving higher-order judgment fragmented across individual courses. Institutions also collect large amounts of activity data without an explicit model linking those data to how judgment develops across a program.

FIRST PROTOTYPE:
Build a program-level capability map for judgment, then use AI to generate case variations, surface evidence gaps, and help faculty compare where learners practice and demonstrate the capability across courses. The analytics layer reports evidence against the model rather than generic engagement.

REPRESENTATIVE USE CASE:
A professional master's program defines a target capability such as 'choose among competing interventions under uncertainty' and traces where learners encounter, practice, and demonstrate it from entry to capstone.

LEARNING / HUMAN-SYSTEM FRAME:
The design makes transfer and progressively complex performance the organizing principle. AI supports case variation and analysis, while faculty define the capability, standards, and acceptable evidence.

QUESTIONS ALREADY IDENTIFIED:
1. Can faculty agree on a small set of program-level judgment capabilities?
2. Which measures predict later performance?
3. Does the architecture reveal unnecessary content or missing practice opportunities?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to exercise domain judgment across varied and increasingly complex situations, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, faculty, courses, assessments, program outcomes, analytics, AI case generator, and institutional governance. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: program-level performance progression, transfer, assessment alignment, faculty agreement, completion, and data usefulness. A key disconfirming signal is: the institution creates a sophisticated map but courses and assessments continue unchanged.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS
TITLE: Linking Frontline Training to Operational Performance

CAPABILITY GAP:
Organizations often measure training completion separately from the operational outcomes the training is supposed to improve. This makes it hard to know whether a performance problem is caused by knowledge, workflow design, equipment, incentives, staffing, or something else.

FIRST PROTOTYPE:
Create an AI-assisted performance diagnostic that combines a capability model with operational error patterns and supervisor observations. Before recommending training, it asks whether the gap is actually learnable and identifies alternative system levers.

REPRESENTATIVE USE CASE:
A manufacturing line shows repeated setup errors. The tool compares errors with task steps and finds that one issue is knowledge-based, another is caused by an ambiguous interface, and only the first should trigger practice.

LEARNING / HUMAN-SYSTEM FRAME:
The human–learner-system frame resists treating every performance gap as a training problem. When learning is appropriate, practice and feedback are tied directly to the target work; when it is not, the system recommends redesign or escalation.

QUESTIONS ALREADY IDENTIFIED:
1. Can the diagnostic distinguish learning gaps from system failures with acceptable reliability?
2. Which operational measures are fair evidence of capability?
3. Does the prototype reduce unnecessary training and improve actual performance?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to perform a frontline task reliably and diagnose whether failures are learnable or systemic, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: worker, supervisor, workflow, equipment/interface, operating conditions, performance data, and AI diagnostic. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: error rate, rework, training avoided, transfer, supervisor agreement, and false training recommendations. A key disconfirming signal is: the tool labels system problems as learner deficits or optimizes a metric workers cannot fully control.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS
TITLE: A Training Harmonization Decision Assistant

CAPABILITY GAP:
Large health and service organizations want to reduce duplicated training, but units may differ in policy, workflow, equipment, and risk. Standardizing too aggressively can make training inaccurate; localizing everything can create high maintenance cost and inconsistent expectations.

FIRST PROTOTYPE:
Create an AI-assisted harmonization tool that classifies training content into system-wide, role-specific, and local-policy components. It generates shared scenarios with controlled local variants and shows exactly which source rule created each variation.

REPRESENTATIVE USE CASE:
Several outpatient units share a common patient-intake capability but differ in escalation policy. The tool produces one core module plus unit-specific decision branches rather than separate courses.

LEARNING / HUMAN-SYSTEM FRAME:
The design preserves common capability targets while representing contextual variation explicitly. Scenario practice and source-grounded feedback support transfer better than simply distributing harmonized information.

QUESTIONS ALREADY IDENTIFIED:
1. What can safely be standardized?
2. Which local differences materially change performance?
3. Does the model reduce maintenance effort while improving accuracy and transfer?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to perform a common professional task correctly across local policy variations, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: learner, educators, multiple units, policies, workflow differences, AI harmonization tool, and governance owners. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: content duplication, update time, policy accuracy, scenario performance, transfer, and local exception rate. A key disconfirming signal is: maintenance decreases but learners are taught a generic workflow that is wrong in important local contexts.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: Transition Readiness Radar

CAPABILITY GAP:
A system changes faster than the people operating it can recalibrate. The formal rollout may be complete while tacit routines, mental models, job aids, and supervision still reflect the old system.

FIRST PROTOTYPE:
Build an AI-assisted transition mapper that compares old and new workflows, identifies changed human decisions and failure points, and generates a small set of scenario checks for readiness rather than a blanket retraining package.

REPRESENTATIVE USE CASE:
A hospital introduces a redesigned medication-administration interface. The prototype highlights the two steps where the new workflow reverses an old habit and creates short interruption-heavy simulations for those moments.

LEARNING / HUMAN-SYSTEM FRAME:
This is a coupling problem between system requirements and human capability. The learning design should target changed decisions and transfer under realistic constraints, while system owners retain responsibility for interface design, policy, and rollout conditions.

QUESTIONS ALREADY IDENTIFIED:
1. Which behaviors actually changed at the transition?
2. What should be redesigned out of the workflow rather than trained around?
3. What evidence shows readiness under realistic operating conditions?

FIRST-CYCLE FRAMING:
Understand: Name the changed system requirement and the operational impact if the old capability persists.
Map: Map old and new workflows, actors, tools, workarounds, supervision, and transition timing; locate where capability-system coupling changed.
Instrument: Measure performance on changed decisions in representative scenarios, transfer on shift, error recovery, workload, and whether failures trace to training or system design.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: Capability Drift Detector

CAPABILITY GAP:
A capability that was once adequate can degrade gradually as tools, staffing, procedures, or local workarounds change. Because there is no single launch event, organizations often notice the problem only after performance has already drifted.

FIRST PROTOTYPE:
Create an AI-assisted drift detector that combines incident narratives, help requests, workaround patterns, procedure changes, and supervisor observations to flag where the required human capability may no longer match the current system.

REPRESENTATIVE USE CASE:
A maintenance organization has updated components and troubleshooting procedures over several years. The detector finds a growing cluster of escalations around one diagnostic step that veteran staff handle informally but newer staff cannot reconstruct.

LEARNING / HUMAN-SYSTEM FRAME:
The prototype treats performance as a property of a changing sociotechnical system, not as a stable trait of workers. Learning interventions are only one possible response; redesign, documentation, staffing, or policy changes may be the stronger lever.

QUESTIONS ALREADY IDENTIFIED:
1. What signals distinguish capability drift from equipment or process drift?
2. How can tacit workarounds be surfaced without normalizing unsafe practice?
3. What threshold should trigger investigation rather than automatic retraining?

FIRST-CYCLE FRAMING:
Understand: Establish the performance outcome that has drifted and the time window over which the system changed.
Map: Map changes in tools, procedures, staffing, knowledge flow, local adaptations, and performance signals to generate rival explanations.
Instrument: Track representative task performance, escalation patterns, error types, knowledge-transfer gaps, and whether intervention effects persist as the system continues to change.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: Capability Requirements Builder

CAPABILITY GAP:
Many systems specify technology, staffing, and process requirements in detail while leaving the human capability requirement vague: 'trained,' 'experienced,' or 'qualified.' If the capability is not defined at observable task grain, it cannot be designed, developed, or verified well.

FIRST PROTOTYPE:
Build an AI-assisted requirements workshop that turns system outcomes, scenarios, hazards, and role descriptions into candidate observable capabilities, then forces human experts to validate, merge, reject, and prioritize them.

REPRESENTATIVE USE CASE:
A new operations center needs analysts who can recognize when an automated alert is misleading. The prototype converts that need into observable decisions, evidence requirements, escalation actions, and representative test scenarios.

LEARNING / HUMAN-SYSTEM FRAME:
The engineering move is requirements definition, but the object being specified is human performance in context. The prototype should expose assumptions and disagreements rather than manufacture false precision.

QUESTIONS ALREADY IDENTIFIED:
1. What must a person actually do, under what conditions, to satisfy the system requirement?
2. Which capabilities are individual versus team or organizational?
3. What evidence would make the requirement verifiable?

FIRST-CYCLE FRAMING:
Understand: Start from the system impact and failure modes, not from an existing course catalog or competency list.
Map: Trace required decisions and actions across roles, interfaces, constraints, and operating scenarios; separate human, team, and system requirements.
Instrument: Validate candidate requirements with work samples, expert agreement, failure cases, and evidence that the requirement predicts or explains operational performance.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: High-Consequence Practice Orchestrator

CAPABILITY GAP:
A capability can be well understood and still fail to scale. Expert instructors, realistic scenarios, feedback time, equipment access, and safe practice opportunities become bottlenecks as programs grow.

FIRST PROTOTYPE:
Create an AI-assisted practice orchestrator that schedules scenario variation, automates low-risk feedback, routes ambiguous performance to experts, and tracks whether quality changes as the learner population scales.

REPRESENTATIVE USE CASE:
A technical workforce must practice fault isolation across many equipment variants. The system generates bounded scenario variations from expert-approved templates while instructors review edge cases and calibration samples.

LEARNING / HUMAN-SYSTEM FRAME:
Scale is not success if the learning mechanism changes. The design should preserve deliberate practice, feedback quality, authenticity, and human oversight while measuring instructor workload and fidelity at increasing volume.

QUESTIONS ALREADY IDENTIFIED:
1. Which parts of expert instruction can be standardized without losing judgment?
2. Where does scale change the pedagogy?
3. What quality signals should stop expansion?

FIRST-CYCLE FRAMING:
Understand: Define the capability, population, throughput requirement, and quality threshold that must survive scale.
Map: Map instructor capacity, scenario supply, feedback loops, equipment constraints, learner variability, and failure consequences.
Instrument: Track learning and transfer, expert-review burden, scenario diversity, feedback reliability, subgroup outcomes, and degradation in fidelity as volume increases.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: Evidence-to-Impact Mapper

CAPABILITY GAP:
Organizations often know who completed training but cannot tell whether the targeted capability changed, transferred to work, or contributed to the operational outcome that justified the intervention.

FIRST PROTOTYPE:
Build an AI-assisted measurement planner that begins with a system outcome, maps the human capability hypothesized to influence it, proposes observable proximal and transfer measures, and surfaces rival explanations before deployment.

REPRESENTATIVE USE CASE:
A service team wants to reduce repeat customer escalations. Instead of treating course completion as evidence, the tool links a specific diagnostic capability to scored work samples, field observations, and downstream escalation patterns.

LEARNING / HUMAN-SYSTEM FRAME:
Measurement is part of the design, not a post hoc dashboard. The prototype should make causal claims smaller and clearer, distinguish capability evidence from operational outcomes, and preserve uncertainty when attribution is weak.

QUESTIONS ALREADY IDENTIFIED:
1. What evidence would convince us the capability changed?
2. What outcome is close enough to the intervention to attribute reasonably?
3. Which rival explanations would make the apparent effect disappear?

FIRST-CYCLE FRAMING:
Understand: Specify the operational decision the evidence must support and the claim being tested.
Map: Build a causal map linking intervention, learner behavior, capability, context, and system outcome; name rival explanations and measurement threats.
Instrument: Predefine work samples, transfer measures, operational indicators, comparison logic, subgroup checks, and stopping rules for claims the evidence cannot support.

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: LENS Canonical
TITLE: Human-AI Delegation Simulator

CAPABILITY GAP:
When AI enters a workflow, the capability requirement changes: people must decide when to rely, when to verify, when to override, and when to revoke delegation. Accuracy alone does not establish that the human-AI team is safe or effective.

FIRST PROTOTYPE:
Create a scenario simulator in which AI advice varies in confidence, quality, and context. Learners practice delegation, verification, override, and escalation while the system records not only outcomes but the reasoning behind reliance decisions.

REPRESENTATIVE USE CASE:
An analyst receives AI-ranked cases with explanations of uneven quality. Some recommendations are right for the wrong reason; others are uncertain. The learner must decide what to accept, inspect, or escalate before seeing consequences.

LEARNING / HUMAN-SYSTEM FRAME:
Human agency is an engineered property of the team. The design should make authority, uncertainty, and correction visible, and test calibrated reliance rather than simple compliance or distrust.

QUESTIONS ALREADY IDENTIFIED:
1. Which decisions may be delegated and under what conditions?
2. Can the human detect when the AI is outside its competence?
3. Does practice improve calibrated reliance without increasing unnecessary workload?

FIRST-CYCLE FRAMING:
Understand: Define the shared task, stakes, authority boundaries, and failure modes of the human-AI team.
Map: Map information flow, AI capabilities and limits, human expertise, time pressure, accountability, override paths, and recovery mechanisms.
Instrument: Measure decision quality, calibration of reliance, verification behavior, override appropriateness, recovery from AI error, workload, and performance when the AI is unavailable.