# Essential Skills for the AI Era > The 17 human and AI skills that appreciate as machines take over execution — sequenced into one curriculum, with exercises, a self-assessment, and a portfolio you build as you go. For students and professionals. - **Audience:** Students · Early-career professionals · Anyone future-proofing their skills for the AI era - **Level:** Foundational → Intermediate (no prerequisites) - **Duration:** 10 weeks · 2 modules/week · 3–5 h/week - **Modules:** 19 - **Pass mark:** 70% - **Interactive version:** https://ragentic.netlify.app/#/courses/essential-skills **This file is generated from the course data by `scripts/build-notes.mjs`. Edit the course data, not this file.** --- ## Phase 0 — Orient (weeks 1) ### Module 1 — The AI-Era Skills Thesis & Your Baseline **Guiding question:** If AI can do more of the execution every year, which of my skills actually appreciate — and where do I stand on them today? **Outcome:** Understand the argument this course rests on: as machine execution gets cheap, the scarce goods become judgment, direction-setting, and the ability to move people. Map the 17 skills across four families, see why the sequence runs meta-skills → thinking → AI fluency → people → character, and take an honest baseline self-assessment you will re-take at the end. **AI-era lens:** Most "future of work" content either panics or hand-waves. This course takes a narrower, more useful position: some skills genuinely depreciate as AI improves, and some genuinely appreciate — and you can tell which is which by asking whether the skill is about producing output or about deciding, judging, and persuading. Your baseline today is not a grade; it is the before-photo that makes your Week-10 progress provable rather than felt. **Apply-at-work mission — Baseline across all 17 skills:** Rate yourself on all 17 skills using the four-level scale (novice / practising / fluent / could teach it). For each, write one sentence of evidence — a real situation that justifies the rating. Then pick the three skills where improvement would most change your next 12 months, and say why. Date it and seal it. 🎯 Completes Capstone Milestone 1. **Reflection:** Which skill did I rate myself highest on — and would someone who works with me agree? Where did I most want to inflate the score, and what does that tell me? #### Resources - [WEF — Future of Jobs Report 2025](https://www.weforum.org/publications/the-future-of-jobs-report-2025/) — The primary evidence base for this course: which skills employers say are rising, which are declining, and how fast. Read the skills-outlook chapter and the top-skills table rather than the whole report. - Name the skills employers report as fastest-rising - Distinguish skills that are declining from those that are stable - Locate your own current skills on the report's outlook - [Andrew Ng — AI For Everyone](https://www.coursera.org/learn/ai-for-everyone) — The most accessible grounding in what AI can and cannot do, free to audit. You need an accurate picture of the machine before deciding which of your skills it actually threatens. - Describe realistically what current AI can and cannot do - Separate genuine capability from marketing claims - Judge which of your own tasks are automatable today - [Ethan Mollick — Co-Intelligence: Living and Working with AI](https://www.penguinrandomhouse.com/books/741805/co-intelligence-by-ethan-mollick/) — A researcher's practical account of working alongside AI rather than being replaced by it — the framing this whole course builds on. - Frame AI as a collaborator with known strengths and failures - Recognise the "jagged frontier" of uneven AI capability - Adopt a stance toward AI that is neither dismissive nor credulous #### Project — Baseline self-assessment across all 17 skills (Capstone component) Rate yourself on all 17 skills using the four-level scale — novice, practising, fluent, could teach it. For each rating, write one sentence of real evidence: a specific situation that justifies it, not an impression. Then identify the three skills where genuine improvement would most change your next twelve months, and explain why those three. Date the document and seal it. Module 19 reopens it and puts the two side by side. **Deliverable:** skills/m01-baseline.md — the dated 17-skill self-assessment with evidence sentences, plus your three priority skills and the reasoning. First component of your Skills Portfolio. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Complete and honest ratings | 30% | All 17 skills rated on the four-level scale, with visible willingness to record low ratings where they are true. | | Evidence, not impressions | 30% | Each rating is justified by a specific real situation rather than a general feeling about yourself. | | Reasoned priorities | 25% | Three priority skills chosen with an argument tied to your actual next twelve months, not to what sounds impressive. | | Sealed and dated | 15% | The document is dated and set aside deliberately so the Module-19 comparison is genuine rather than reconstructed. | #### Scenario drills **Drill 1.** You ask an AI model to explain a topic you know extremely well, and it produces a confident, polished answer. **Task:** Read it as a critic: list everything it got subtly wrong, oversimplified, or asserted without support. Then ask yourself what a non-expert would have accepted. **Drill 2.** You are tempted to rate yourself "fluent" on a skill because you understand it intellectually. **Task:** Write the last three times you actually used that skill under pressure and what the outcome was. Then re-rate. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The core argument of this course is that as AI improves…** a. Every human skill loses value at roughly the same steady rate b. Some skills depreciate while judgment-heavy ones appreciate c. Only technical skills will end up mattering to employers d. Human skills become entirely irrelevant to the job market **2. A skill is more likely to APPRECIATE in the AI era when it is mainly about…** a. Producing a standard output quickly b. Deciding, judging, or moving people c. Remembering large amounts of information d. Following an established procedure exactly **3. Why does this course put metacognition and learnability first?** a. They are by far the easiest of the seventeen to pick up quickly b. They are the meta-layer every later skill compounds off c. They are the two skills employers ask about most often in interviews d. They happen to appear first in the source research **4. Critical thinking is placed BEFORE the AI-technical modules because…** a. Prompting is too advanced to start with b. You must be able to judge the output before you delegate the work c. Critical thinking is the more important skill d. The technical modules take far more time **5. The baseline self-assessment exists so that…** a. The course has an assignment to grade b. Your Week-10 progress is provable rather than merely felt c. Your tutor can decide what you skip d. You can compare scores against other learners on the course **6. The four skill families in this course are…** a. Cognitive, interpersonal, character, and AI fluency b. Technical, managerial, financial, and creative ability c. Reading, writing, arithmetic, and coding d. Junior, intermediate, senior, and executive career levels **7. Writing evidence sentences beside each self-rating matters because…** a. Longer documents are graded more favourably b. It forces honesty by demanding a specific real situation as proof c. The template will not accept ratings without them d. It makes the assessment take longer **8. The most useful stance toward AI that this course argues for is…** a. Treating it as an infallible oracle you should defer to b. A capable collaborator with known failure modes c. A passing fad best ignored until it settles down d. A direct threat to be resisted wherever possible **9. If you rate yourself "could teach it" on a skill, that should mean…** a. You feel broadly confident about the skill b. You could get another person to real competence in it yourself c. You completed a formal course on it d. You have read a good deal about it **10. Choosing three priority skills rather than working on all seventeen equally reflects…** a. A limitation of how the course happens to be structured b. That focused improvement beats diffuse effort c. That the other fourteen skills matter considerably less d. That most learners will not finish the whole programme anyway --- ## Phase 1 — Think — Cognitive Skills (weeks 1–3) ### Module 2 — Metacognition: Thinking About Your Thinking **Guiding question:** How do I notice when I am reasoning deliberately versus running on autopilot? **Outcome:** Build the executive function that governs every other cognitive skill: notice your default patterns, journal decisions with reasoning and confidence, identify your three most persistent biases and build tripwires for them, and run pre-mortems and the Feynman technique to expose gaps before they cost you. **AI-era lens:** Metacognition comes first because it is the skill that makes every later skill improvable. It also becomes sharply more valuable with AI: when a fluent, confident answer is always one prompt away, the deciding question is whether you noticed yourself accepting it without thinking. The learner who can catch their own autopilot is the one who uses AI as leverage rather than as a substitute for judgment. **Apply-at-work mission — Decision journal + bias tripwires:** Start a decision journal: for five real decisions this week record the decision, your reasoning, your confidence as a percentage, and what would prove you wrong. Then name your three most persistent cognitive biases and write a specific tripwire for each — the observable moment that should make you stop and check. **Reflection:** Where this week did I notice myself defaulting to habit rather than reasoning? What did catching it change? #### Resources - [Coursera — Mindshift: Break Through Obstacles to Learning](https://www.coursera.org/learn/mindshift) — Barbara Oakley's follow-up to Learning How to Learn, free to audit. The most accessible route into noticing and changing your own thinking patterns. - Notice your default thinking patterns as they happen - Identify where habit is substituting for deliberate reasoning - Apply mindset-shift techniques to a real stuck problem - [Farnam Street — Mental Models & Second-Order Thinking](https://fs.blog/mental-models/) — A practical, free library of thinking tools. Read selectively: inversion, second-order thinking, and the map-is-not-the-territory entries repay the most attention. - Apply inversion and second-order thinking to a decision - Build a small personal toolkit of mental models - Recognise when a model is being over-applied - [Donella Meadows — Thinking in Systems: A Primer](https://www.chelseagreen.com/product/thinking-in-systems/) — The advanced text: how to see structures and feedback loops rather than isolated events. Slow reading; unusually high return for leaders and analysts. - See feedback loops behind recurring problems - Distinguish events, patterns and underlying structure - Find leverage points rather than treating symptoms #### Project — Decision journal + bias tripwires (Portfolio component) Start a decision journal and run it for a week on five real decisions. For each, record the decision, your actual reasoning, your confidence as a percentage, and what evidence would prove you wrong. Then identify the three cognitive biases you fall for most often and write a specific tripwire for each — the observable moment that should stop you and make you check. Set a date to revisit the predictions. **Deliverable:** skills/m02-decision-journal.md — five journalled decisions with reasoning and confidence, three named biases with tripwires, and a revisit date. Portfolio evidence for the cognitive family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Real decisions, real reasoning | 30% | Five genuine decisions with the reasoning you actually used, not a tidied-up version written afterwards. | | Confidence stated numerically | 25% | Each decision carries a confidence percentage and a falsifier — what would show you were wrong. | | Specific, observable tripwires | 30% | Three biases named with tripwires you could actually notice in the moment, not general intentions to be careful. | | Revisit scheduled | 15% | A dated point to check predictions against outcomes — calibration needs the loop closed. | #### Scenario drills **Drill 1.** You have a decision to make and you notice you already know what you want to do. **Task:** Before acting, ask an AI model to argue the strongest case AGAINST your preferred option, then write which of its points you cannot answer. **Drill 2.** You want to test whether you actually understand something you have just learned. **Task:** Explain it to an AI model as if teaching a twelve-year-old, then ask it to identify what your explanation left out or got wrong. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Metacognition is best described as…** a. Thinking rather faster than you currently do b. Monitoring and regulating your own thinking c. A meditation practice used for improving your focus d. A memory technique used for revising before exams **2. Recording your confidence as a percentage in a decision journal lets you…** a. Justify the decision later to other people b. Track calibration — whether your confidence matches reality c. Make decisions substantially more quickly than before d. Avoid having to explain your reasoning at all **3. A pre-mortem works by…** a. Reviewing what went wrong after the project has already finished b. Assuming failure has happened, then working backwards to why c. Asking your line manager to approve the whole plan in advance d. Listing every possible risk in a large spreadsheet **4. The Feynman technique exposes gaps in understanding by requiring you to…** a. Read through the source material several more times b. Explain the concept simply, as if to a beginner c. Test yourself repeatedly using a deck of flashcards d. Discuss the topic at length with an expert **5. A useful bias tripwire is one that is…** a. Ambitious in its scope and fairly broadly worded b. Specific and observable in the moment c. Reviewed once a year during your annual appraisal d. Shared publicly with your whole immediate team **6. Metacognition becomes MORE valuable in the AI era mainly because…** a. AI systems can now do the reflecting on your behalf b. A fluent answer is always one prompt away, so noticing your own acceptance matters c. Employers have started testing for it in interviews d. It is one of the easier skills to demonstrate quickly **7. Dialectical journaling means writing…** a. A daily record of the events exactly as they happened b. The arguments for and against a belief you hold c. Your main goals for the coming twelve months d. A summary of what you read that week **8. Rumination differs from reflection in that rumination…** a. Takes place in writing rather than in your head b. Loops without producing a decision or change c. Focuses on other people rather than yourself d. Happens immediately after an event occurs **9. Tracking prediction against outcome over time builds…** a. A defensible audit trail for your manager b. Calibration — a realistic sense of your own accuracy c. A faster decision-making process day to day d. Evidence that you were right more often than not **10. The strongest sign your metacognition is improving is that you…** a. Feel considerably more confident in your decisions b. Catch yourself mid-autopilot more often than before c. Make decisions in noticeably less time than you used to d. Are able to explain your reasoning to other people ### Module 3 — Learnability: Re-Skilling as a Compounding Asset **Guiding question:** How do I learn genuinely new things fast enough to keep up — and make that ability compound? **Outcome:** Design a personal learning system: understand how memory and recall actually work, use spaced repetition to retain rather than cram, build a personal knowledge management system you will actually maintain, set deliberate stretch challenges, and test real understanding by teaching what you learned within 48 hours. **AI-era lens:** The half-life of specific technical knowledge keeps shortening while the value of being able to acquire it keeps rising. AI accelerates this in both directions: it makes any subject explainable on demand, which is a gift to a learner with a system and a trap for one without — endless fluent explanations feel like learning while nothing is retained. Learnability is what converts AI's explaining power into durable capability. **Apply-at-work mission — Your personal learning system:** Build and document a learning system you will actually use: a capture tool (Obsidian, Notion, or paper), a spaced-repetition habit, and a review rhythm. Then run it on something real — learn one genuinely new concept this week and teach it to another person within 48 hours. Record what the teaching attempt exposed. **Reflection:** What did trying to teach it reveal that reading it never would have? Where did my understanding turn out to be borrowed rather than owned? #### Resources - [Coursera — Learning How to Learn](https://www.coursera.org/learn/learning-how-to-learn) — The most-enrolled MOOC ever built, free to audit. Focused and diffuse modes, chunking, procrastination and memory — the mechanics behind everything else in this module. - Explain how focused and diffuse thinking each contribute to learning - Use chunking and spaced retrieval to retain material durably - Apply a concrete technique against your own procrastination pattern - [Nicky Case — How to Remember Anything Forever-ish](https://ncase.me/remember/) — A free, playable explainer on spaced repetition. Twenty minutes that will change how you handle anything you need to actually retain. - Describe why spaced repetition beats re-reading - Set up a spaced-repetition habit you will maintain - Choose what genuinely deserves memorising - [Anders Ericsson — Peak: Secrets from the New Science of Expertise](https://www.hachettebookgroup.com/titles/anders-ericsson/peak/9780544456259/) — The advanced read: what deliberate practice actually requires, and why most practice does not qualify. The research behind expertise, from its principal researcher. - Distinguish deliberate practice from ordinary repetition - Design practice at the edge of your competence - Build the feedback loop deliberate practice requires #### Project — Your personal learning system (Portfolio component) Design and document a learning system you will genuinely maintain: a capture tool, a spaced-repetition habit, a review rhythm, and a rule for what deserves to be captured at all. Then run it on something real — learn one genuinely new concept this week using the system, and teach it to another person within 48 hours. Record specifically what the teaching attempt exposed about the gaps in your understanding. **Deliverable:** skills/m03-learning-system.md — the documented system, the concept you learned, and an honest account of what teaching it revealed. Portfolio evidence for the cognitive family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | A system you will actually keep | 30% | Realistic tools and rhythm matched to your actual life, not an aspirational setup you will abandon in a fortnight. | | Run on something real | 25% | Applied to a genuinely new concept this week rather than described in the abstract. | | Taught within 48 hours | 25% | Actually explained to another person, with their questions recorded. | | Honest gap analysis | 20% | Names what the teaching exposed — the borrowed understanding — rather than reporting that it went well. | #### Scenario drills **Drill 1.** You have just read something complex and you are not sure how much of it stuck. **Task:** Close the source. Explain it to an AI model from memory, then ask it to grade your explanation against the actual concept and list what you missed. **Drill 2.** You want to learn a genuinely new domain and do not know where to start. **Task:** Ask an AI model to build you a syllabus with a logical order, then ask it what a beginner most commonly misunderstands at each stage. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Spaced repetition works better than re-reading because it…** a. It takes considerably less total time overall to do b. Forces effortful recall just as forgetting begins c. Covers a greater volume of the material per session d. Feels rather more comfortable and less frustrating **2. Teaching something within 48 hours of learning it is valuable because…** a. It helps the other person more than it helps you b. Explaining exposes the gaps that reading conceals c. It creates a useful record for your own notes d. Most courses require some form of assessment **3. Deliberate practice differs from ordinary repetition in that it…** a. Involves considerably more hours of work each week b. Sits at the edge of ability, with feedback c. Is always supervised by a qualified coach d. Focuses on the parts you are already good at **4. A personal knowledge management system fails most often because…** a. The chosen software lacks important features b. It is more elaborate than the owner will maintain c. There is not enough material worth capturing d. The material captured is not organised alphabetically **5. Learnability matters more in the AI era mainly because…** a. AI has now made most existing learning materials entirely obsolete b. Specific knowledge dates faster while acquiring it compounds c. Employers now require a formal certificate for every single skill d. Learning has become significantly easier than it once was **6. The particular trap AI creates for learners is that…** a. It refuses to explain genuinely difficult concepts b. Fluent explanation on demand feels like learning c. It is only accurate on very simple subjects d. It cannot produce examples or analogies well **7. A good "stretch challenge" is one that is…** a. Comfortably within your existing ability b. Just beyond your current competence c. Far outside anything you have attempted d. Identical to what your peers are doing **8. Tracking your learning velocity over time tells you…** a. Which of the subjects you happen to enjoy the most b. How long you now take to reach competence c. How you compare against the other learners d. Which of the learning materials are best written **9. Cross-domain learning is valuable primarily because…** a. It looks rather impressive on a professional CV b. Insight transfers between fields that rarely meet c. It is much easier than deep specialisation d. It requires far less sustained effort to maintain **10. The most reliable sign you have truly learned something is that you can…** a. Recognise it immediately when you see it again b. Explain it clearly and use it in a new situation c. Recall the specific source where you first read it d. Feel confident about it when the topic comes up ### Module 4 — Critical Thinking: Judging What You Are Told **Guiding question:** How do I question assumptions and evaluate claims — including the confident ones a machine hands me? **Outcome:** Build rigorous reasoning: question assumptions before accepting conclusions, use Socratic questioning to get beneath first answers, recognise logical fallacies in real time including your own, steelman opposing views, and apply first-principles and second-order thinking to real decisions. **AI-era lens:** The World Economic Forum ranks analytical and critical thinking at the very top of employer-sought skills — and AI raises the stakes rather than lowering them. A model produces fluent, well-formatted, confident text whether or not it is right, which means the burden of judgment shifts entirely to you. This module deliberately comes BEFORE the AI-technical modules: you learn to judge before you learn to delegate. 🎯 Completes Capstone Milestone 2. **Apply-at-work mission — Steelman + assumption audit:** Take a position you hold strongly and steelman the opposing view — write the strongest honest case against yourself, then respond to it. Separately, take one AI-generated answer on a topic you know well, and audit it: list every assumption it made, every claim you cannot verify, and every place it was confidently wrong or subtly incomplete. **Reflection:** What did steelmanning change about my own position — did it soften, sharpen, or survive intact? What did the AI get wrong that a non-expert would have accepted? #### Resources - [Coursera — Critical Thinking Skills for University Success](https://www.coursera.org/learn/critical-thinking-skills) — University of Sydney, free to audit. The cleanest entry point: how to read a claim, separate evidence from assertion, and build an argument that survives scrutiny. - Separate a claim from the evidence offered for it - Question assumptions before accepting a conclusion - Construct an argument that survives a hostile read - [Coursera — Introduction to Logic and Critical Thinking (Duke)](https://www.coursera.org/specializations/logic-critical-thinking-duke) — Walter Sinnott-Armstrong's "Think Again" specialisation, free to audit. Formal enough to be rigorous, informal enough to use on Monday. - Identify the structure of an argument in ordinary prose - Name common formal and informal fallacies on sight - Judge whether a conclusion actually follows - [Daniel Kahneman — Thinking, Fast and Slow](https://us.macmillan.com/books/9780374533557/thinkingfastandslow) — The advanced read: the Nobel-winning account of the two systems and the biases that live in the fast one. Slow going, permanently useful. - Recognise fast-system judgments as they happen - Name the biases you personally fall for most - Design decisions that do not rely on willpower #### Project — Steelman + AI assumption audit (Portfolio component) Two halves. First, take a position you hold strongly and write the strongest honest case against it — the version its best advocate would recognise as fair — then write your response. Second, generate an AI answer on a topic you know well and audit it line by line: every assumption it made, every claim you cannot verify, and every place it was confidently wrong or quietly incomplete. Finish with a note on what a non-expert would have accepted. **Deliverable:** skills/m04-critical-thinking.md — a steelman with your response, and a marked-up AI answer with an assumption/verification audit. Portfolio evidence for the cognitive family. 🎯 Completes Capstone Milestone 2. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | The steelman is genuinely strong | 30% | The opposing case is one its best advocate would accept as fair, not a version built to be knocked over. | | Your response engages the strong version | 20% | You answer the steelman as written, and say plainly where you cannot. | | Assumptions surfaced in the AI answer | 25% | Unstated assumptions named specifically, not a general observation that the model made some. | | Verification separated from plausibility | 25% | Each claim marked verified, unverifiable, or wrong — with the check you actually ran, not how right it sounded. | #### Scenario drills **Drill 1.** You have an AI answer that looks complete and you are about to use it. **Task:** Ask the model to list every assumption its own answer depends on, then ask a second model to find the weakest one. Verify that assumption yourself before proceeding. **Drill 2.** You are about to commit to a position in a report or a meeting. **Task:** Ask an AI model to write the strongest rebuttal a well-informed critic would give, then write down which of its points you cannot answer. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Steelmanning an opposing argument means…** a. Restating it in its strongest and most honest form b. Finding its weakest version and attacking it c. Agreeing with it publicly to avoid a conflict d. Listing the factual errors the other side made **2. Socratic questioning is mainly a technique for…** a. Getting beneath a first answer to its reasoning b. Winning an argument in front of an audience c. Gathering facts quickly from a subject expert d. Testing how confident a speaker actually feels **3. Second-order thinking asks…** a. And then what happens after that? b. Who was responsible for this decision? c. How quickly can this be delivered now? d. What did the last team try in this case? **4. First-principles reasoning means…** a. Stripping a problem back to what you know is true b. Copying the approach the market leader has taken c. Starting from the first idea that comes to your mind d. Following the standard process used in that industry **5. Critical thinking sits BEFORE the AI modules in this course because…** a. You must judge output before you delegate work b. Prompting is technically harder than reasoning is c. Employers test reasoning before AI tool use d. The WEF ranks it first among all skills **6. The most dangerous property of AI-generated text for a critical thinker is that it…** a. Reads as fluent and confident whether or not it is right b. Is usually far too long to read carefully end to end c. Refuses to answer questions on genuinely contested topics d. Contains obvious errors that almost any reader would catch **7. Attacking the person making an argument rather than the argument itself is…** a. Ad hominem — the arguer, not the argument b. Straw man — a weakened version of it c. False dilemma — only two options given d. Circular reasoning — assuming the conclusion **8. An assumption audit of an AI answer should list…** a. What it took for granted and what you cannot verify b. Every sentence that contains a factual-sounding claim c. The model version and the generation date d. The sources it cites, ranked by their apparent credibility **9. Confirmation bias is best countered by…** a. Seeking the evidence that would prove you wrong b. Gathering more evidence supporting your position c. Waiting longer before committing to a conclusion d. Asking a colleague whether they agree with your view **10. A claim you cannot verify should be treated as…** a. Unverified, and flagged as such in your output b. Probably true if the rest of the answer checked out c. False until some credible source has confirmed it d. Unimportant if it does not change the conclusion ### Module 5 — Financial Literacy: Economic Self-Defence **Guiding question:** Do I actually understand compounding, debt, and how a business makes money — well enough to make good decisions? **Outcome:** Build practical financial capability: understand compounding as both an asset and a cost, read a balance sheet, P&L and cash-flow statement, track your own net worth, use ROI and payback thinking on decisions, and understand the tax-advantaged structures available in your jurisdiction. **AI-era lens:** Financial literacy is included here because it is the most consistently neglected compounding skill, and AI changes little about it — the arithmetic was always available, yet the knowledge gap persists and tracks directly to long-run wealth inequality. What AI does change is access to explanation: you can now interrogate a financial statement or a pension rule conversationally. That only helps if you know which questions matter, which is what this module builds. **Apply-at-work mission — Read one real annual report:** Pick a company you know and read one annual report cover to cover. Write a one-page summary in plain language: how does this business actually make money, what does it own and owe, where is the cash going, and what would worry you as an owner? Separately, calculate your own net worth and set a monthly tracking habit. **Reflection:** What surprised me about how this business really makes money versus how I assumed it did? What financial decision am I currently making on vibes rather than arithmetic? #### Resources - [Coursera — Financial Planning for Young Adults](https://www.coursera.org/learn/financial-planning) — University of Illinois, free to audit. Budgeting, credit, debt and investing basics without jargon — the module to take first if any of this feels opaque. - Explain how compounding works for saving and for debt - Build a budget and a net-worth tracker you will keep - Describe the main types of consumer credit and their costs - [Coursera — Financial Markets (Yale, Robert Shiller)](https://www.coursera.org/learn/financial-markets-global) — A Nobel laureate on how markets, risk and institutions actually work. Free to audit. The best available bridge from personal finance to macro understanding. - Explain risk, diversification and the role of insurance - Describe how markets price assets and where that fails - Connect macro forces to your own financial decisions - [Benjamin Graham — The Intelligent Investor](https://www.harpercollins.com/products/the-intelligent-investor-rev-ed-benjamin-graham) — The advanced classic on temperament, margin of safety, and the difference between investing and speculating. Read the commentary edition. - Distinguish investing from speculation honestly - Apply margin-of-safety thinking to a real decision - Recognise the behavioural errors that cost most returns #### Project — Annual report read-through + personal net worth (Portfolio component) Pick a company whose product you actually use and read its most recent annual report end to end. Write a one-page plain-language summary: how it really makes money, what it owns and owes, where the cash went, and the one thing that would worry you as an owner. Then, separately and privately, calculate your own net worth, set up a monthly tracking habit, and name one decision you are currently making on instinct where the arithmetic is available. **Deliverable:** skills/m05-financial-literacy.md — the one-page company summary plus your tracking method and the decision you re-did with arithmetic. Keep the net-worth figures private if you prefer; the method is the evidence. Portfolio evidence for the cognitive family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Plain-language business model | 30% | You can say how the money is made in ordinary words, without repeating the report's own phrasing. | | All three statements used | 25% | Balance sheet, P&L and cash flow each contribute something to the summary rather than one carrying it all. | | One genuine owner's worry | 20% | A specific concern grounded in a number you found, not a generic risk anyone could list. | | A tracking habit that will survive | 25% | A concrete monthly method with a trigger, plus one real decision redone on arithmetic instead of instinct. | #### Scenario drills **Drill 1.** You are looking at a financial statement or a pension rule you do not fully understand. **Task:** Paste the section into an AI model and ask it to explain it as if to a smart beginner, then ask what question an expert would ask about it that you have not thought to ask. **Drill 2.** You are about to make a financial decision on instinct — a subscription, a loan, a purchase. **Task:** Ask an AI model to build the arithmetic with you: total cost over the full term, the alternative use of the money, and the break-even point. Then decide. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Compounding is described here as economic self-defence because it…** a. Works just as powerfully against you in debt b. Guarantees a return on any long-term investment c. Protects your savings from inflation completely d. Is the fastest available route to serious wealth **2. A cash-flow statement tells you…** a. Where money actually moved in and out b. What the business is worth on the market c. How profitable each product line has been d. What the company owns and what it owes **3. The balance sheet shows…** a. What a business owns and what it owes b. Revenue and costs over a trading period c. Cash generated by operating activities d. Dividends paid to shareholders in a year **4. Payback period measures…** a. How long until an investment repays its cost b. The total profit the investment will produce c. The interest rate implied by a set of flows d. How much of a loan still remains outstanding **5. Tracking your net worth monthly is useful mainly because…** a. You can only manage what you actually measure b. It tells you precisely when you can retire safely c. Lenders require a monthly statement of net assets d. It removes the need to budget month by month **6. What AI genuinely changes about financial literacy is that it…** a. Makes explanation cheap, not judgment about what matters b. Can now manage a portfolio better than any human adviser c. Removes any need to understand compounding arithmetic d. Has made most personal-finance qualifications redundant **7. Reading one annual report cover to cover is valuable because…** a. It shows how a real business explains its own money b. Annual reports are always written in plain language c. It is the fastest way to value a company correctly d. The figures are audited, so they cannot be misleading **8. Return on investment compares…** a. The gain from a decision against its cost b. This year's performance against last year c. One company against its closest competitor d. Total revenue against the whole market size **9. A tax-advantaged account (ISA, pension, 401(k)) mainly helps by…** a. Letting returns compound without annual tax drag b. Guaranteeing a higher rate of return than cash does c. Protecting the money from any market downturn d. Allowing you to withdraw the funds at any moment **10. The financial habit this module most wants to break is…** a. Deciding on instinct where arithmetic was available b. Choosing investments that are far too cautious c. Failing to negotiate hard enough on your salary d. Holding too much of your wealth as plain cash ### Module 6 — Storytelling: Making Ideas Land **Guiding question:** How do I make an idea land — so it is remembered, believed, and acted on? **Outcome:** Build narrative capability: master the setup → conflict → resolution structure, find the emotional hook before reaching for data, tailor to the audience's world rather than your own, reverse-engineer great talks, practise micro-storytelling daily, and build a personal story bank you can deploy in any context. **AI-era lens:** Narrative is retained far better than raw facts, and this gap widens as AI makes facts abundant. When anyone can generate a competent summary, the scarce skill is deciding what the point is and making people care about it. AI is genuinely useful here — as a drafting partner and a ruthless first critic — but it cannot supply the lived specificity that makes a story land, and audiences increasingly detect the difference. **Apply-at-work mission — Build your story bank:** Write ten short personal anecdotes — real situations from your study or work — each with a clear setup, tension, and insight. Tag each with the point it can make. Then take your most recent piece of dry work (a report, a slide, an update) and rewrite its opening as a story that earns the reader's attention in 90 seconds. **Reflection:** Which of my stories actually moved someone when I told it, and which fell flat? What did the difference come down to? #### Resources - [Coursera — Storytelling and Influencing: Communicate with Impact](https://www.coursera.org/learn/storytelling-influencing) — Macquarie University, free to audit. Practical structure for business narrative: audience, arc, and the move from informing to persuading. - Structure a message as setup, conflict and resolution - Identify what a specific audience actually cares about - Convert a dry update into something worth listening to - [Coursera — Dynamic Public Speaking (University of Washington)](https://www.coursera.org/specializations/public-speaking) — Free to audit. Delivery as a trainable craft — how to build, rehearse and land a talk rather than reading a document aloud. - Design a talk around one clear central idea - Rehearse deliberately rather than re-reading slides - Use pace, pause and emphasis to carry meaning - [Robert McKee — Story: Substance, Structure, Style](https://www.harpercollins.com/products/story-robert-mckee) — The advanced text, written for screenwriters and read by everyone who has to hold an audience. Structure at a level most business communication never reaches. - Analyse why a scene creates or releases tension - Build an arc that earns its ending - Diagnose why a story you told fell flat #### Project — Story bank + one dry thing rewritten (Portfolio component) Write ten short anecdotes from your own study or work — real situations, each with a clear setup, a genuine tension, and the insight it earns. Tag each with the point it can be used to make. Then take your most recent piece of dry work — a report, a slide, a status update — and rewrite its opening so it earns the reader's attention within ninety seconds. Keep the original alongside it so the difference is visible. **Deliverable:** skills/m06-story-bank.md — ten tagged anecdotes plus a before-and-after rewrite of one real piece of your own work. Portfolio evidence for the cognitive and interpersonal families. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Ten real, specific anecdotes | 30% | Drawn from things that actually happened to you, with the concrete detail that makes them yours. | | Each has genuine tension | 25% | Something was at stake and could have gone otherwise — not a smooth account of a task completed. | | Tagged to a point | 15% | Each story is labelled with the argument it can carry, so you can find it when you need it. | | The rewrite earns attention | 30% | Before and after shown together, and the after gives a reader a reason to continue in the first few lines. | #### Scenario drills **Drill 1.** You have written something true, useful, and completely flat. **Task:** Give it to an AI model and ask it to identify where the tension is, and where a reader would stop caring. Do not ask it to rewrite the piece — rewrite the opening yourself from what it found. **Drill 2.** You have a real anecdote but you are not sure what point it makes. **Task:** Tell it to an AI model in full and ask what three different arguments it could support, and what detail each version would need to lead with. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The core story structure used in this module is…** a. Setup, then conflict, then resolution b. Claim, then evidence, then conclusion c. Problem, then options, then decision d. Context, then data, then a request **2. Leading with the emotional hook matters because…** a. Data convinces, but emotion moves people to act b. Audiences cannot follow numeric arguments c. Emotional stories are quicker to write than reports d. Facts on their own are always misremembered **3. Tailoring a story to your audience means starting from…** a. Their goals and their world, not yours b. The most impressive result you achieved c. The chronological order of what happened d. The detail that took you the most work **4. A story bank is…** a. A library of real anecdotes tagged by the point they make b. A folder of slide templates you reuse across your own decks c. A shared archive of your organisation's customer testimonials d. A collection of quotations from the writers you most admire **5. Micro-storytelling practice means…** a. Telling one 90-second story about a real situation daily b. Writing a brief summary of every meeting that you attend c. Cutting every presentation you give down to three slides d. Posting short updates about your work each morning **6. Storytelling appreciates in the AI era because…** a. When facts are abundant, deciding the point is scarce b. AI cannot write grammatically correct narrative prose c. Audiences now read considerably less than they used to d. Story structure is difficult to teach a language model **7. The real limit of AI as a storytelling partner is that it…** a. Cannot supply the lived specificity that makes it land b. Is unable to suggest any usable narrative structure at all c. Produces prose that is grammatically wrong d. Will not write in the first person **8. Reverse-engineering a great talk means…** a. Mapping its arc to see how it was built b. Memorising the delivery and imitating it c. Summarising its argument in a short note d. Fact-checking every claim the speaker made **9. Naming the story in a data presentation means…** a. Saying explicitly what the chart is telling you b. Giving each chart a short descriptive title c. Adding a caption with the data source below d. Putting the headline number in a larger font **10. The clearest evidence your story actually worked is that…** a. Someone retells it later, to somebody else b. The audience said it was very interesting c. You got through it without losing your place d. People asked questions about the details --- ## Phase 2 — Direct AI — Technical Fluency (weeks 4) ### Module 7 — Prompting & Vibe Coding: Directing the Machine **Guiding question:** How do I direct an AI model deliberately instead of hoping a vague request works? **Outcome:** Build AI interaction fluency: supply role, goal, constraints, audience and output format every time; treat the first prompt as a hypothesis and iterate; evaluate outputs critically because models are confidently wrong; use chain-of-thought and few-shot patterns; build a personal prompt library; and understand different models' strengths and failure modes. **AI-era lens:** McKinsey found AI-fluency demand in job postings rose sharply between 2022 and 2024, and prompt-related requirements now appear across finance, law, healthcare and product — this is becoming a baseline literacy like spreadsheets. But notice the sequence: you arrive here having already built metacognition and critical thinking, because prompting without judgment produces fluent nonsense faster. The skill is not typing clever prompts; it is specifying well and evaluating honestly. **Apply-at-work mission — Your prompt library:** Build a documented prompt library for your own recurring work: at least eight prompts that reliably work, each with the pattern it uses (role, few-shot, chain-of-thought, critique), an example output, and a field note on where it fails. Test at least two of them across different models and record the differences. **Reflection:** Which of my prompts failed in ways I did not expect? Where did a better brief fix what I first blamed on the model? #### Resources - [DeepLearning.AI — ChatGPT Prompt Engineering for Developers](https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/) — Free short course from Andrew Ng and Isa Fulford. About ninety minutes, and the fastest route from ad-hoc prompting to deliberate prompting. Not as developer-only as the title suggests. - Write prompts with explicit role, task and format - Use iteration as a method rather than as retrying - Recognise where a model will confidently fail - [Coursera — Prompt Engineering for ChatGPT (Vanderbilt)](https://www.coursera.org/learn/prompt-engineering) — Jules White's course on prompt patterns — persona, template, flipped interaction, chain of thought. Paid, but the pattern vocabulary is what makes prompting teachable rather than folkloric. - Apply named prompt patterns to your own recurring tasks - Combine patterns rather than writing every prompt fresh - Design a prompt that asks the model to interview you - [DeepLearning.AI — Building Systems with the ChatGPT API](https://www.deeplearning.ai/short-courses/building-systems-with-chatgpt/) — The advanced option: chaining calls, evaluating outputs, and building something that behaves reliably rather than impressively once. Free. - Break a task into chained, checkable steps - Build an evaluation set before trusting a pipeline - Handle the failure cases rather than the demo case #### Project — Your prompt library (Portfolio component) Build a documented prompt library for work you actually do repeatedly. At least eight prompts, each with: the pattern it uses, the prompt itself, an example of good output, and a field note on where it fails or degrades. Test at least two of the prompts across two different models and record the differences you observed rather than the differences you expected. Finish with the one prompt you tried to build and abandoned, and why. **Deliverable:** skills/m07-prompt-library.md — eight or more documented prompts with patterns, examples and failure notes, plus a cross-model comparison. Portfolio evidence for the AI fluency family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Eight prompts for real recurring work | 25% | Drawn from tasks you genuinely repeat, not generic prompts copied from a list online. | | Pattern named for each | 20% | Each prompt is labelled with the pattern it uses, so the library teaches you patterns and not just prompts. | | Failure notes present and specific | 30% | Each entry records where the prompt breaks down — the input that defeats it, the output you had to reject. | | Cross-model comparison observed | 25% | Two prompts run on two models with the actual differences recorded, including where there was no difference worth noting. | #### Scenario drills **Drill 1.** You have a task in mind but you are not sure how to brief it properly. **Task:** Use the flipped interaction: "Before you answer, ask me the questions you need answered to do this well." Answer them, then let it proceed. **Drill 2.** A model has produced an answer that looks good and you are inclined to accept it. **Task:** Start a fresh conversation, paste the output without saying where it came from, and ask a model to critique it against your rubric. Compare with your own read. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The single highest-value thing to add to a weak prompt is usually…** a. Context: role, goal, constraints, audience, format b. A politer and more conversational opening line c. An instruction to think carefully and be accurate d. A request that the model avoid making mistakes **2. Treating your first prompt as a hypothesis means…** a. You expect to revise it based on what comes back b. You should write it as a question, never a command c. You test it on several models before you rely on it d. You keep it short until you know it works well **3. Few-shot prompting means…** a. Giving examples of the output you want b. Asking for only a few options in the reply c. Keeping the whole prompt to a few sentences d. Running the same prompt a few times **4. Chain-of-thought prompting helps most on tasks that…** a. Require several dependent reasoning steps b. Need a creative or unusual turn of phrase c. Involve summarising a long piece of text d. Have one short and factual right answer **5. The "flipped interaction" pattern involves…** a. Asking the model to interview you before it answers b. Answering the model's first draft with your own critique c. Giving the model a persona opposed to your own d. Reversing the order of the instructions in your prompt **6. Testing a prompt across different models is worth doing because…** a. Each has different strengths and failure modes b. The cheapest model is always the right choice c. Outputs should match if the prompt is good d. It proves the prompt has no ambiguity **7. This module comes AFTER critical thinking in the course because…** a. Prompting without judgment just produces fluent nonsense faster b. Prompt patterns are conceptually harder than logical fallacies c. Most employers assess reasoning ability before AI tool fluency d. The technical modules require the earlier cognitive ones as prerequisites **8. A good field note in a prompt library records…** a. The input that defeats the prompt b. How many tokens the prompt consumes c. The date the prompt was first written d. Which model you first tested it on **9. The claim that AI fluency demand rose sharply in job postings comes from…** a. McKinsey's analysis of postings between 2022 and 2024 b. A World Economic Forum survey of graduate recruiters c. Coursera enrolment figures for prompt engineering courses d. OpenAI's published statistics on enterprise API usage **10. The real skill this module builds is…** a. Specifying well and evaluating honestly b. Memorising a large set of prompt templates c. Knowing which model leads the benchmarks d. Writing prompts that are as short as possible ### Module 8 — AI Workflow Integration: From User to Orchestrator **Guiding question:** How do I rebuild my actual working week around AI instead of using it occasionally? **Outcome:** Move from ad-hoc use to designed workflow: audit your work for repetitive, pattern-based and research-heavy tasks; integrate deeply into one workflow before scaling; write clear AI briefs; build evaluation rubrics so you know when output meets your standard; and define the human/AI division of labour explicitly. **AI-era lens:** The competitive edge belongs not to AI itself but to the person who combines judgment with AI leverage — the role shifts from task-executor to orchestrator who sets goals, reviews outputs and applies judgment. For a student this is the difference between using AI to finish an assignment and using it to learn faster than peers; for a professional it is the difference between saving minutes and redesigning a role. 🎯 Completes Capstone Milestone 3. **Apply-at-work mission — 30-day workflow integration:** Audit your week and pick the single most time-consuming recurring task. Design an AI-integrated version of it: what the AI does, what you do, what you check, and how you will know the output is good enough. Run it for a fortnight, measure the before and after honestly, and write up what actually changed — including what got worse. **Reflection:** Where did AI genuinely save me time, and where did it just move the effort into reviewing? What did I stop being able to do well because I stopped doing it? #### Resources - [Coursera — AI For Everyone (DeepLearning.AI, Andrew Ng)](https://www.coursera.org/learn/ai-for-everyone) — The best non-technical introduction to what AI can and cannot do, free to audit. Written for people who have to make decisions about AI rather than build it. - Describe realistically what current AI can and cannot do - Identify which tasks in a workflow are AI-suitable - Talk about AI projects without technical hand-waving - [Coursera — Generative AI for Everyone (DeepLearning.AI)](https://www.coursera.org/learn/generative-ai-for-everyone) — The follow-up, focused specifically on generative tools in daily work: where they fit, how to evaluate them, and how teams adopt them without chaos. Free to audit. - Map generative tools onto your own weekly tasks - Judge when a task is worth automating at all - Plan an adoption that survives contact with a team - [Ethan Mollick — Co-Intelligence: Living and Working with AI](https://www.penguinrandomhouse.com/books/741805/co-intelligence-by-ethan-mollick/) — The advanced read: how work actually changes when a capable, unreliable collaborator joins it. Empirical, sceptical, and more useful than most strategy writing on the subject. - Apply the "always invite AI to the table" heuristic critically - Recognise the jagged frontier of model capability - Design human-in-the-loop work that stays honest #### Project — 30-day workflow integration, measured (Portfolio component) Audit your week and pick the single most time-consuming recurring task. Design an AI-integrated version: what the AI does, what you do, what you check, and the rubric that decides whether the output is good enough. Run it for at least a fortnight. Then write it up honestly — the before and after in real numbers where you have them, what improved, what got worse, and what you noticed yourself getting slower at because you stopped doing it by hand. **Deliverable:** skills/m08-workflow-integration.md — the workflow design, the evaluation rubric, before-and-after measurements, and an honest account of the costs. Portfolio evidence for the AI fluency family. 🎯 Completes Capstone Milestone 3. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | One real workflow, chosen by audit | 20% | Selected because the audit showed it costs the most time, not because it was the easiest to automate. | | Division of labour stated explicitly | 25% | It is clear what the AI does, what you do, and what you check — with the checkpoint written down, not implied. | | Evaluation rubric defined before running | 25% | A written standard for "good enough" that existed before the results did, so it could not be adjusted to fit them. | | Honest measurement, including the losses | 30% | Real before-and-after figures plus what got worse — review burden, skill atrophy, quality drift on the hard cases. | #### Scenario drills **Drill 1.** You suspect a recurring task could be AI-integrated but you are not sure which part. **Task:** Describe the task to an AI model step by step and ask it to split the steps into three groups: safe to delegate, needs your review, must stay human. Then argue with its answer. **Drill 2.** You need a standard for whether AI output on your task is good enough. **Task:** Take three past outputs — one you accepted, one you rejected, one you were unsure about — and ask a model to infer the rubric that distinguishes them. Edit it into your own. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The audit step comes first because…** a. You should automate what costs the most time b. The easiest task is the safest place to start c. Managers usually require an audit beforehand d. Every workflow needs documenting eventually **2. Depth of integration beats breadth of adoption because…** a. One workflow rebuilt properly teaches you more than ten touched lightly b. Using several tools at once is prohibited by most organisations c. Deep integration is considerably cheaper than broad integration d. It is generally easier to learn one tool than to learn several **3. The shift from task-executor to orchestrator means your main work becomes…** a. Setting goals, reviewing output, applying judgment b. Writing longer and more elaborate prompts each time c. Managing the tool subscriptions your team relies on d. Producing more of the same output in less time **4. An evaluation rubric for AI output should be written…** a. Before you see any results b. After the first week of running c. Once the workflow has stabilised d. Only if quality problems appear **5. The most commonly missed cost of AI integration is that…** a. Time saved on producing gets spent on reviewing b. The subscription fees exceed the productivity gains c. Colleagues resent the change and stop cooperating d. Output volume rises faster than anyone can use it **6. Skill atrophy in this context means…** a. Losing capability at things you stopped doing yourself b. Forgetting how to use the AI tool after a break c. Team members becoming less willing to learn tools d. Prompts becoming less effective as models change **7. The "170 million new jobs" figure in the source deck comes from…** a. World Economic Forum projections on AI augmentation b. McKinsey's analysis of job postings across sectors c. An OECD study of automation in member countries d. Andrew Ng's own estimates in the AI For Everyone course **8. A clear AI brief resembles…** a. A brief you would give a capable new colleague b. A precise technical specification with no ambiguity c. A short instruction that trusts the model to infer d. A list of the constraints the output must not break **9. Documenting your team's human/AI division of labour mainly helps by…** a. Making the checkpoints explicit rather than assumed b. Reducing the total number of licences the team needs c. Proving to leadership that the team is using AI well d. Ensuring everybody uses exactly the same prompts **10. The honest measure of a successful integration is…** a. A real before-and-after, including what got worse b. A strong feeling that the week runs more smoothly c. The number of tasks now handled by the AI d. How quickly the whole team adopted the workflow --- ## Phase 3 — Connect — Interpersonal Skills (weeks 5–7) ### Module 9 — Human Behavior: What Actually Moves People **Guiding question:** What actually drives the people around me — and how do I read it accurately? **Outcome:** Build people-reading capability: study motivation frameworks including self-determination theory, observe before diagnosing, distinguish stated from revealed preferences, understand the principles of influence and social proof, and use 360-degree feedback to see yourself as others do. **AI-era lens:** Interpersonal awareness is among the most AI-resistant capabilities in McKinsey's analysis, because understanding motivation requires reading context, history and unspoken incentive — the things not written down. As AI handles more analysis, the bottleneck in most work moves to getting humans aligned and moving. AI can help you rehearse and prepare for a conversation; it cannot notice that your colleague has gone quiet in meetings since the reorg. **Apply-at-work mission — Motivation map + observation log:** Pick a real team, class group, or project you are part of. Spend a week observing before diagnosing: log what people actually do, not what they say. Then map each person's apparent motivation against a framework (self-determination theory works well) and identify one place where stated and revealed preferences diverge — and what that implies about how to work with them. **Reflection:** Where was I wrong about what motivated someone? What did I learn from watching rather than asking? #### Resources - [Coursera — Inspiring and Motivating Individuals (Michigan)](https://www.coursera.org/learn/motivate-people-teams) — Free to audit. Maxim Sytch on what actually drives performance: needs, goals, rewards, and the traps managers fall into. Practical rather than inspirational. - Diagnose a motivation problem rather than label a person - Distinguish intrinsic from extrinsic drivers in real cases - Spot rewards that are quietly pulling against the goal - [Coursera — Organizational Analysis (Stanford)](https://www.coursera.org/learn/organizational-analysis) — Free to audit. Several competing lenses for reading why organisations behave as they do — the antidote to explaining everything through personality. - Apply more than one lens to the same situation - Explain behaviour by structure, not just character - Predict how a decision will actually be received - [Robert Cialdini — Influence: The Psychology of Persuasion](https://www.harpercollins.com/products/influence-new-and-expanded-robert-b-cialdini) — The advanced read on the six principles and why they work. Read it as defence as much as offence — you will notice how often they are being used on you. - Name the principle at work in a real persuasion attempt - Recognise when a principle is being used on you - Use influence ethically, or decline to use it #### Project — Motivation map + observation log (Portfolio component) Pick a real group you belong to — a team, a class cohort, a project. Spend a week observing before you diagnose anything: log what people actually do, not what they say, and keep the log factual. Then map each person's apparent motivation against a framework — self-determination theory (autonomy, competence, relatedness) works well — and find one place where a stated preference and a revealed preference diverge. Write what that divergence implies about how to work with them, and what you might have got wrong. **Deliverable:** skills/m09-motivation-map.md — a week of observational notes, a motivation map for the group, and one documented stated-versus-revealed divergence with its implications. Portfolio evidence for the interpersonal family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Observation precedes diagnosis | 30% | A week of factual notes on what people did, written before any conclusions about why. | | A framework actually applied | 25% | Motivations mapped against a named framework rather than described in loose adjectives. | | A real stated-versus-revealed gap | 25% | One concrete divergence, evidenced from your log, not a general observation that people are inconsistent. | | Held loosely | 20% | You name what you might be wrong about and what evidence would change the map. | #### Scenario drills **Drill 1.** You are about to conclude that a colleague is being difficult or unreasonable. **Task:** Describe only their observable behaviour to an AI model — no adjectives, no interpretation — and ask for five different explanations, ranked by how likely each is. **Drill 2.** You need to understand what actually motivates someone you work with closely. **Task:** Write out what they say they want and what their behaviour suggests they want, then ask an AI model where those two accounts contradict each other. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The core of self-determination theory is that people are driven by…** a. Autonomy, competence and relatedness b. Recognition, security and advancement c. Achievement, affiliation and power d. Fairness, clarity and predictability **2. Revealed preferences are…** a. What behaviour shows people actually want b. What people say they want when you ask them c. What the organisation formally rewards people for d. What a personality assessment predicts **3. Observing before diagnosing matters because…** a. Early diagnosis freezes an explanation before you have evidence b. Watching quietly is faster than interviewing everybody c. Managers are formally required to document team issues d. It gives people time to correct the behaviour themselves **4. A motivation problem is more usefully treated as…** a. A condition of the work rather than a flaw in the person b. A recruitment error that should be corrected as quickly as possible c. A sign that the compensation package needs to be revised upward d. Evidence that the individual is in the wrong career **5. Rewarding A while hoping for B means…** a. The incentives pull against the stated goal b. The goal has been set much too far in the future c. The reward is not large enough to matter d. The team does not understand the strategy **6. Reading people is AI-resistant mainly because it requires…** a. Context, history and unspoken incentive that are not written down b. A level of emotional processing no model has yet achieved c. Direct physical presence in the room with the other person d. Access to private information the model cannot legally see **7. Useful AI support in this module looks like…** a. Rehearsing and preparing for a conversation b. Diagnosing a colleague from your description c. Drafting the message you will send afterwards d. Predicting how the person will likely respond **8. The main value of 360-degree feedback comes from…** a. The patterns across responses, not single comments b. The average score taken across all of the categories c. The comments coming from the most senior respondents d. The gap between your self-score and the mean **9. Social proof describes the tendency to…** a. Take behavioural cues from what others are doing b. Trust people who hold recognised formal authority c. Reciprocate favours that have been done for you d. Stay consistent with commitments made publicly **10. Culture, in the sense used in this module, is…** a. Behaviour at scale b. A published set of values c. The tone leadership sets d. How people feel at work ### Module 10 — Empathy & Compassion: Understanding Before Responding **Guiding question:** How do I genuinely understand another person's experience rather than waiting for my turn to speak? **Outcome:** Build empathic capability: practise active listening to understand rather than to reply, make perspective-taking a deliberate habit, recognise emotional contagion and your own effect on a room, use mindfulness to reduce the reactivity that blocks empathy, and apply empathy mapping to build shared understanding of others' experience. **AI-era lens:** Social-emotional capability is consistently identified as among the hardest to replicate with AI, and its value rises as more interaction becomes machine-mediated. There is a real and current danger worth naming: AI can generate text that sounds empathetic, which tempts people to outsource the appearance of care while skipping the substance. Recipients notice. Genuine empathy remains the human differentiator precisely because it cannot be convincingly faked at scale. **Apply-at-work mission — Deep listening practice:** Have three conversations this week where your only goal is to understand — no advising, no fixing, no waiting to speak. Afterwards write what you learned that you would have missed, and what you were tempted to say instead. Then complete an empathy map for one person whose behaviour you find difficult, from their point of view. **Reflection:** What did listening properly reveal that I had assumed wrongly? Where did I feel the pull to fix rather than understand? #### Resources - [Coursera — The Science of Well-Being (Yale, Laurie Santos)](https://www.coursera.org/learn/the-science-of-well-being) — Yale's most popular course ever, with over five million enrolled and free to audit. Evidence-based rather than inspirational, and the sections on social connection and kindness carry this module. - Explain why our predictions about what helps are unreliable - Practise kindness and connection as tested interventions - Build one well-being habit that survives past week two - [Coursera — Foundations of Positive Psychology (UPenn)](https://www.coursera.org/specializations/positivepsychology) — Martin Seligman's specialisation, free to audit. Deeper on the research base: strengths, resilience, and what actually improves relationships. - Describe the research base behind positive interventions - Identify your own signature strengths honestly - Apply an intervention and measure whether it worked - [Brené Brown — Atlas of the Heart](https://www.penguinrandomhouse.com/books/659976/atlas-of-the-heart-by-brene-brown/) — The advanced read: eighty-seven emotions mapped precisely. The value is granularity — you cannot respond well to a feeling you can only call "bad". - Name emotions with more precision than good or bad - Distinguish empathy from sympathy in practice - Recognise the emotion behind a difficult reaction #### Project — Deep listening practice + empathy map (Portfolio component) Have three conversations this week where your only goal is to understand — no advising, no fixing, no waiting for your turn. Afterwards, write what you learned that you would otherwise have missed, and what you were tempted to say instead of listening. Then build an empathy map for one person whose behaviour you find genuinely difficult: what they see, hear, say, do, think and feel, written from their point of view rather than yours. **Deliverable:** skills/m10-empathy-practice.md — three listening write-ups and a complete empathy map for one difficult relationship, written in their voice. Portfolio evidence for the interpersonal family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Three real conversations, understanding only | 30% | Genuine conversations where you did not advise or fix, with what you learned recorded afterwards. | | The temptation named | 20% | You record what you wanted to say instead of listening — the honest part, and the most useful. | | Empathy map written from their side | 30% | Their perspective in their terms, including the parts that make their behaviour reasonable to them. | | A changed assumption | 20% | At least one belief you held about someone that the listening actually altered. | #### Scenario drills **Drill 1.** You are dreading a conversation with someone whose position you find unreasonable. **Task:** Ask an AI model to role-play that person, briefed with their situation and constraints rather than your judgment of them. Have the conversation, then ask it what you missed. **Drill 2.** You want to check whether a message you have written actually sounds like it comes from a person who cares. **Task:** Ask a model to read it as the recipient would, given their situation, and say what it would feel like to receive. Then rewrite it yourself. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Active listening means listening…** a. To understand rather than to reply b. Without interrupting the other person c. While taking notes on what is said d. And then summarising it back to them **2. Emotional contagion describes how…** a. Your emotional state spreads to those around you b. Strong feelings intensify when they are suppressed c. Groups reach agreement faster under time pressure d. Anxiety increases as a deadline draws closer **3. Empathy differs from sympathy in that empathy…** a. Enters the other person's perspective rather than viewing it b. Is felt more strongly and expressed more openly c. Applies to strangers as well as to close friends d. Requires you to have had the same experience **4. Emotional granularity means…** a. Naming what you feel precisely rather than vaguely b. Feeling emotions much more intensely than most people c. Keeping a daily record of your emotional states d. Separating your emotions from your decisions **5. The AI-era danger this module names explicitly is that…** a. Text can sound empathetic while the care is absent b. Models refuse to engage with emotional subjects c. People will stop having difficult conversations at all d. AI advice on relationships is frequently inaccurate **6. Perspective-taking as a deliberate habit means asking…** a. Why might a reasonable person do this? b. What would I have done if I were in their place? c. How would I feel if this happened to me? d. What is the kindest reading available here? **7. An empathy map is most useful when it is written…** a. In the other person's voice, not yours b. After a long and honest conversation c. Together with the person it describes d. With the whole team contributing to it **8. Mindfulness supports empathy chiefly by…** a. Reducing the reactivity that blocks it under pressure b. Making you more aware of other people's feelings c. Creating a pause before you have to speak at all d. Lowering the general stress level across your working day **9. Social-emotional capability is considered AI-resistant because…** a. It rests on presence and history, not on producing text b. Models are not trained on emotional conversation data c. It requires a physical, in-person interaction to work d. Emotional intelligence cannot be measured objectively **10. The strongest sign your listening actually worked is that…** a. You learned something that changed your view b. The other person spoke for most of the time available c. You resisted every urge to offer them any advice d. The conversation lasted much longer than expected ### Module 11 — Feedback Fluency: The Highest-Leverage Management Tool **Guiding question:** How do I give feedback that actually helps — and receive it without defending? **Outcome:** Build feedback capability both ways: use the situation-behaviour-impact model to stay specific and observable, separate intent from impact, give feedback close to the event, ask for feedback with a specific question rather than a vague invitation, and model receiving it without defensiveness so others can be honest with you. **AI-era lens:** Feedback is the highest-leverage tool available for improving anyone's performance, including your own — and poor feedback is a leading reason strong performers disengage. AI can now review your work instantly and tirelessly, which is genuinely useful for drafts and code, but it cannot tell you the thing your manager notices and will not say. Learning to extract honest human feedback is a skill that AI abundance makes rarer and more valuable. **Apply-at-work mission — Give one hard piece of feedback, ask for one:** Give one piece of genuinely useful feedback using situation-behaviour-impact — something you have been avoiding, delivered with care and specificity. Separately, ask one person for feedback with a specific question ("what is one thing I do in meetings that gets in the way?"). Write up both: what you said, how it landed, and what you learned. **Reflection:** What made the feedback I gave land or fail? How did I react internally to what I was told — and what does that reaction reveal? #### Resources - [Coursera — Giving Helpful Feedback (University of Colorado)](https://www.coursera.org/learn/feedback) — Free to audit. Short, practical, and centred on the situation-behaviour-impact model. The fastest way to stop giving feedback that is technically true and completely useless. - Structure feedback as situation, behaviour and impact - Keep the observation factual and free of interpretation - Deliver it close enough to the event to be usable - [Coursera — Inspiring and Motivating Individuals (Michigan)](https://www.coursera.org/learn/motivate-people-teams) — Free to audit. The final module on feedback and performance appraisal is the relevant one here — how feedback interacts with rewards, and why appraisal so often fails. - See how feedback and reward systems interact - Seek feedback rather than waiting to receive it - Give critical feedback without destroying motivation - [Douglas Stone & Sheila Heen — Thanks for the Feedback](https://www.penguinrandomhouse.com/books/313485/thanks-for-the-feedback-by-douglas-stone-and-sheila-heen/) — The advanced read, and unusually about RECEIVING feedback — the harder and more neglected half. From the authors of Difficult Conversations. - Identify your own trigger pattern when criticised - Separate the truth question from the relationship one - Ask for feedback in a way that gets you honesty #### Project — Give one hard piece of feedback, ask for one (Portfolio component) Two exchanges, both real. First, give one genuinely useful piece of feedback you have been avoiding, using situation-behaviour-impact — written out before you deliver it, and delivered with care. Second, ask one person for feedback using a specific question rather than a general invitation: "what is one thing I do in meetings that gets in the way?" Write up both: what you prepared, what you actually said, how it landed, and — for the one you received — what you noticed yourself feeling before you responded. **Deliverable:** skills/m11-feedback-exchange.md — one SBI script with an account of how it landed, and one received-feedback write-up including your internal reaction. Portfolio evidence for the interpersonal family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | SBI used properly | 25% | A specific situation, observable behaviour, and stated impact — no character judgments dressed as observations. | | It was actually difficult | 20% | Feedback you had been avoiding, not a safe piece chosen because it was easy to deliver. | | The request was specific | 25% | One narrow question that made an honest answer easy, rather than "any feedback for me?" | | Your own reaction recorded honestly | 30% | What you felt on hearing it, including defensiveness — the reaction is the data this exercise exists to surface. | #### Scenario drills **Drill 1.** You have feedback to give and you keep postponing the conversation. **Task:** Write your SBI in full, then ask an AI model to identify every place you have slipped a judgment in as an observation. Rewrite until only behaviour remains. **Drill 2.** You have received feedback that stung and you want to respond well rather than fast. **Task:** Write down what you were told and your honest first reaction. Ask a model to separate the substantive claim from the delivery, then decide what you accept. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. The SBI model stands for…** a. Situation, behaviour, impact b. Specific, balanced, immediate c. Setting, background, insight d. Support, behaviour, improve **2. Separating intent from impact means recognising that…** a. Meaning well does not change what the action did b. Intent matters more than the outcome achieved c. Impact should be discussed only in private d. Most workplace harm is entirely accidental **3. Feedback should be given close to the event because…** a. Detail and relevance both decay quickly b. Delayed feedback is considered unprofessional c. It fits better into a weekly meeting rhythm d. People forget the feedback if you wait too long **4. Asking "any feedback for me?" tends to fail because…** a. It is easier to answer with a reassuring nothing b. Most people are not qualified to assess you c. It puts the other person on the spot socially d. The answer is usually far too general to use **5. Care personally and challenge directly means…** a. Doing both at once, not choosing between them b. Building the relationship before any challenge c. Being direct with peers and gentle with juniors d. Alternating supportive and critical conversations **6. What AI genuinely cannot do for you here is…** a. Tell you the thing your manager notices and will not say b. Review a draft or a piece of code against clearly stated criteria c. Help you rehearse a difficult conversation well in advance d. Suggest more precise wording for an observation **7. Modelling how to receive feedback matters because…** a. People calibrate honesty to how the last person reacted b. It demonstrates that you are open to being challenged c. Receiving well is harder than giving feedback well d. It sets the tone for the rest of the conversation **8. A feedback statement that says "you were unprofessional" fails mainly because…** a. It is a judgment, not an observable behaviour b. It is too harsh to be received without defence c. It should have been delivered in writing first d. It does not offer a suggested improvement **9. Poor feedback is described as a leading reason that…** a. Strong performers disengage and leave b. Teams miss their quarterly objectives c. Managers are moved out of their roles d. New hires fail their probation periods **10. Your own defensiveness when receiving feedback is best treated as…** a. Information about where you are exposed b. A normal reaction that is best ignored c. A sign the feedback was poorly delivered d. Evidence that the criticism was unjustified ### Module 12 — Emotional Vulnerability: Going First **Guiding question:** How do I share real uncertainty in a way that builds trust rather than undermining confidence? **Outcome:** Build appropriate openness: distinguish vulnerability (authentic, purposeful sharing) from emotional dumping, start small with trusted people before scaling to groups, name emotions precisely, and understand that a team cannot be open if the person leading it never goes first. **AI-era lens:** A decade of research on vulnerability finds it is the origin of connection, creativity and trust rather than a weakness — and it is a strategic capability, not a personality quirk. Its value grows in an AI era for a specific reason: when polished output is free, authenticity becomes the scarce signal. "I don't know yet" from a real person carries information that no amount of confident generated prose can substitute for. **Apply-at-work mission — Go first, once, deliberately:** Identify one genuine uncertainty or mistake you have been hiding at work or in your studies, and share it appropriately with someone whose trust matters — not as confession, but as honest information. Note what changed in the relationship. Then write your own line between vulnerability and oversharing, with two examples on each side of it. **Reflection:** What did I fear would happen if I admitted this, and what actually happened? Where am I still performing certainty I do not have? #### Resources - [Brené Brown — The Power of Vulnerability (TED)](https://www.ted.com/talks/brene_brown_the_power_of_vulnerability) — Twenty minutes, free, and the clearest statement of the core research finding: vulnerability is the origin of connection, not a weakness to be managed. - State the research case for vulnerability plainly - Distinguish vulnerability from disclosure for its own sake - Identify where you are performing certainty - [Amy Edmondson — Building a Psychologically Safe Workplace (TEDx)](https://www.youtube.com/watch?v=LhoLuui9gX8) — Free. The organisational half of the argument: psychological safety as a measurable team property, from the researcher who defined it. Her seven-item scale is worth finding. - Define psychological safety precisely, not loosely - Assess a team you belong to against the concept - See why safety and high standards are compatible - [Brené Brown — Daring Greatly](https://www.penguinrandomhouse.com/books/204669/daring-greatly-by-brene-brown/) — The advanced read: the full research case for vulnerability as a leadership asset, including the shame dynamics that make people avoid it. - Recognise the defences that block honest sharing - Set your own boundary between openness and dumping - Go first, deliberately, in a group that needs it #### Project — Go first, once, deliberately (Portfolio component) Identify one genuine uncertainty, mistake, or gap you have been concealing at work or in your studies. Share it appropriately with one person whose trust matters — not as a confession, but as honest and useful information. Record what you feared would happen and what actually happened. Then write your own boundary between vulnerability and oversharing, with two worked examples on each side of the line, in your own context rather than in the abstract. **Deliverable:** skills/m12-going-first.md — the account of what you shared, the fear versus the outcome, and your own stated boundary with four worked examples. Portfolio evidence for the interpersonal and character families. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Something real was shared | 30% | A genuine uncertainty or mistake you had been concealing, not a safe admission chosen for the exercise. | | Fear and outcome both recorded | 25% | What you expected to happen sits beside what actually happened, and the gap is named. | | A boundary you can actually use | 25% | A stated line between vulnerability and dumping, in your own words and your own context. | | Four worked examples | 20% | Two on each side of the line, specific enough that someone else could apply your rule. | #### Scenario drills **Drill 1.** You are unsure whether what you want to share is vulnerability or oversharing. **Task:** Describe the situation and what you plan to say to an AI model, and ask it to argue both readings — first that it serves the relationship, then that it burdens the other person. **Drill 2.** You have been performing certainty about something you are not certain about. **Task:** Write the honest version of what you know and do not know, then ask a model to help you phrase it so it stays useful rather than alarming. Say it to a person, not the model. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Vulnerability differs from emotional dumping mainly in that vulnerability is…** a. Purposeful, and considers the other person b. Shared with fewer people at any one time c. Expressed more calmly and less intensely d. Confined to things that are already resolved **2. Psychological safety means a team where people believe…** a. They can raise a problem without being punished b. Their job is secure regardless of their performance c. Conflict between colleagues is actively avoided d. Standards are relaxed enough to reduce the stress **3. Leaders going first matters because…** a. Nobody risks more exposure than the person above them b. It is the fastest way to build rapport with a new team c. People generally imitate whatever behaviour they observe d. It demonstrates that the leader trusts the team already **4. Vulnerability becomes MORE valuable in the AI era because…** a. When polished output is free, authenticity is the scarce signal b. Machine-mediated interaction has replaced most conversation c. Models are unable to generate text expressing uncertainty d. Remote work has removed the informal moments that built trust **5. Emotional granularity contributes here by…** a. Making a feeling specific enough to act on b. Reducing the intensity of difficult emotions c. Helping others understand you more quickly d. Preventing you from overreacting to setbacks **6. Starting small means…** a. One real uncertainty, one trusted person b. Sharing only things of little consequence c. Waiting until the relationship is established d. Rehearsing what you will say beforehand **7. Edmondson's seven-item scale is useful because it…** a. Turns a felt sense into something you can track b. Diagnoses which team member is causing problems c. Provides a benchmark against industry averages d. Replaces the need for one-to-one conversations **8. "I don't know yet" from a leader tends to…** a. Build more trust than confident guessing does b. Undermine confidence in their overall decision-making c. Invite the team to challenge every decision d. Signal that the leader is out of their depth **9. The commonest misreading of vulnerability at work is treating it as…** a. A personality trait rather than a practice b. A tactic for getting other people to like you c. Something only senior people can afford d. An excuse for lowering the work standard **10. The honest test of whether your sharing was appropriate is…** a. Whether it gave the other person something useful b. Whether you felt considerably better about it afterwards c. Whether the other person responded warmly to what you said d. Whether you would say the same thing in a public setting ### Module 13 — Conflict Resolution: Disagreeing Productively **Guiding question:** How do I handle a real disagreement so the relationship and the outcome both survive? **Outcome:** Build negotiation and conflict capability: distinguish positions from underlying interests, manage your own emotional state before engaging, separate the people from the problem, know your default conflict style, and use interest-based approaches to find trades that a positional fight would never reach. **AI-era lens:** Conflict resolution is rated highly AI-resistant because it demands contextual empathy, creative trade-offs, and reading what is not being said. It is also the skill most people avoid practising, which is exactly why it differentiates. AI has a real supporting role — rehearsing a difficult conversation with a model before having it for real is genuinely useful preparation — but the conversation itself is irreducibly human. 🎯 Completes Capstone Milestone 4. **Apply-at-work mission — Positions vs interests, on a live disagreement:** Take a real current disagreement. Write both sides' stated positions, then dig for the underlying interests behind each. Identify one trade that serves both sets of interests and would never have emerged from arguing positions. If appropriate, have the conversation — rehearse it with an AI partner first — and record what actually happened. **Reflection:** What interest was hiding behind the other person's position? Where did my own emotional state shape the outcome more than my argument did? #### Resources - [Coursera — Negotiation Fundamentals (ESSEC Business School)](https://www.coursera.org/learn/negotiation-fundamentals) — Free to audit. The clean introduction: preparation, the negotiation sequence, and why most people concede value they never needed to concede. - Prepare a negotiation before entering the room - Separate the issues from the relationship - Recognise the standard concession traps - [Coursera — Successful Negotiation: Essential Strategies and Skills (Michigan)](https://www.coursera.org/learn/negotiation-skills) — Free to audit. Principled negotiation in full: BATNA, reservation price, and the zone of possible agreement, with practice cases. - Calculate your BATNA before you begin negotiating - Find the zone of possible agreement in a real case - Trade across issues instead of splitting one - [Chris Voss — Never Split the Difference](https://www.harpercollins.com/products/never-split-the-difference-chris-vosstahl-raz) — The advanced read: tactical empathy, labelling, and calibrated questions, from an FBI hostage negotiator. Read alongside the principled-negotiation material rather than instead of it. - Use labelling to defuse an emotional exchange - Ask calibrated questions instead of making demands - Recognise these techniques when used on you #### Project — Positions versus interests on a live disagreement (Portfolio component) Take a real disagreement you are currently in. Write both sides' stated positions plainly, then dig underneath each for the interests driving them — what each side actually needs, which is rarely what they are asking for. Identify at least one trade that serves both sets of interests and could never have emerged from arguing positions. If it is appropriate, have the conversation — rehearse it with an AI partner first — and record what actually happened, including where your prediction was wrong. **Deliverable:** skills/m13-positions-interests.md — both positions, the interests beneath them, at least one interest-based trade, and an account of the conversation if you had it. Portfolio evidence for the interpersonal family. 🎯 Completes Capstone Milestone 4. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Both positions stated fairly | 20% | Theirs written in terms they would recognise and accept as accurate. | | Interests genuinely excavated | 30% | The needs underneath both positions, including at least one that surprised you. | | A trade that positions would never reach | 30% | A specific exchange serving both sets of interests, not a compromise splitting the difference. | | Own emotional state accounted for | 20% | You name how your own state shaped the exchange — before, during, or in what you avoided. | #### Scenario drills **Drill 1.** You have a difficult conversation coming and you keep rehearsing it in your head, badly. **Task:** Brief an AI model with the other person's situation, constraints and likely interests — not your grievance — and role-play the conversation twice: once where they are reasonable, once where they are not. **Drill 2.** You are stuck arguing over a single number or a single decision. **Task:** List everything either side cares about, then ask an AI model to propose five trades across those issues that neither of you has raised. Discard the unworkable ones yourself. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. A position differs from an interest in that a position is…** a. What someone says they want b. What someone will settle for c. What someone can justify publicly d. What someone asks for at the start **2. Your BATNA is…** a. Your best alternative if there is no agreement b. The lowest offer that you are willing to accept c. The most you could plausibly ask for here d. The outcome both parties would prefer most **3. Separating the people from the problem means…** a. Both of you against the problem, not each other b. Discussing the issue without emotion involved c. Removing personal relationships from decisions d. Handing the disagreement to a neutral party **4. Managing your own emotional state first matters because…** a. Threat responses narrow the options you can see b. Showing emotion weakens your negotiating position c. The other party will mirror whatever you display d. Calm people are perceived as more trustworthy **5. The zone of possible agreement is…** a. The overlap between what each side would accept b. The range of offers made during the whole discussion c. The midpoint between the two opening positions taken d. The set of issues both parties have agreed to discuss **6. Labelling, in Voss's sense, means…** a. Naming the emotion you observe in the other person b. Categorising the type of conflict you are in c. Giving each issue a name to track it clearly d. Stating your own position in explicit terms **7. Conflict resolution is rated highly AI-resistant because it needs…** a. Contextual empathy and reading what is unsaid b. Faster processing than a person can manage alone c. Legal knowledge that models frequently get wrong d. Physical presence for the agreement to be binding **8. The legitimate AI role in this module is…** a. Rehearsal beforehand, not the conversation itself b. Drafting the message that finally resolves the dispute c. Judging which party has the stronger position d. Mediating between both parties in real time **9. Knowing your default conflict style is useful because…** a. Your default fires before you have chosen anything b. Some styles are simply more effective than all the others c. It tells you which conflicts you should avoid d. It explains why other people find you difficult **10. An interest-based trade is better than splitting the difference because…** a. Each side gives up what it values least b. It reaches an agreement more quickly c. It avoids the need to discuss the numbers d. It feels fairer to both parties involved --- ## Phase 4 — Lead — Character Traits (weeks 7–9) ### Module 14 — Self-Reflection: Honest Self-Assessment **Guiding question:** Can I see myself accurately — including the gap between my stated values and my actual behaviour? **Outcome:** Build accurate self-perception: distinguish reflection that moves you forward from rumination that loops, ask what you contributed before assessing others, confront the gap between stated values and actual behaviour, find a genuine truth-teller, and build annual personal reviews into your calendar. **AI-era lens:** Growth-mindset research is clear that honest self-assessment — as distinct from self-criticism or victim narrative — is what separates people who keep improving from those who plateau. In an AI era this matters more, not less: when the machine can do more each year, the person who accurately knows what they are good at, and honestly tracks what they are losing, adapts. The person running on a flattering self-image gets surprised. **Apply-at-work mission — Values-behaviour gap audit:** Write your five stated core values. For each, find real evidence from the last three months of how you actually behaved — calendar, messages, decisions. Name the widest gap between stated and lived, without excusing it, and one specific change that would close it. Then ask a truth-teller whether they recognise your self-assessment. **Reflection:** Which gap was hardest to write down honestly? Did my truth-teller see me the way I see myself — and what did the difference teach me? #### Resources - [Coursera — Everyday Leadership: The Art of Leading Others (Duke)](https://www.coursera.org/learn/everyday-leadership) — Free to audit. Leadership as ordinary practice rather than position — and unusually direct about honest self-assessment as its precondition. - Separate leadership behaviour from formal authority - Assess your own contribution to an outcome first - Identify the habits you would need to change - [Carol Dweck — The Power of Believing That You Can Improve (TED)](https://www.ted.com/talks/carol_dweck_the_power_of_believing_that_you_can_improve) — Ten minutes, free. The growth-mindset research stated plainly, and the distinction that matters here: honest assessment is not self-criticism. - State what growth mindset does and does not claim - Distinguish honest assessment from self-criticism - Reframe a current failure as information - [Tasha Eurich — Insight: The Surprising Truth About How Others See Us](https://www.penguinrandomhouse.com/books/536063/insight-by-tasha-eurich/) — The advanced read. Her research finds most people believe they are self-aware and most are not — and that introspection alone rarely fixes it. The chapters on truth-tellers are the practical core. - Distinguish internal from external self-awareness - Recognise why introspection can mislead you - Recruit and use a genuine truth-teller #### Project — Values-behaviour gap audit (Portfolio component) Write down your five stated core values. For each, go looking for real evidence from the last three months of how you actually behaved — your calendar, your messages, the decisions you made, the things you declined. Name the widest gap between stated and lived without excusing it, and one specific change that would close it. Then show your self-assessment to a truth-teller and ask whether they recognise the person you described. Record where they disagreed. **Deliverable:** skills/m14-values-gap-audit.md — five values with behavioural evidence, the widest gap named plainly, one closing change, and your truth-teller's response. Portfolio evidence for the character family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Evidence, not intention | 30% | Each value tested against real artifacts — calendar, decisions, messages — rather than how you generally try to behave. | | The widest gap named without excuse | 25% | Stated plainly, with no justifying context added to soften it. | | One specific closing change | 20% | A concrete behavioural change with a trigger, not a resolution to do better. | | Truth-teller consulted | 25% | A real person read your assessment, and where they disagreed is recorded rather than argued away. | #### Scenario drills **Drill 1.** You have written a self-assessment and you suspect it is flattering. **Task:** Give it to an AI model and ask it to write the version a well-informed critic who dislikes you would write from the same facts. Then mark which of its points you cannot honestly dismiss. **Drill 2.** You keep returning to the same setback without getting anywhere. **Task:** Ask a model to help you convert the loop into three specific questions with answerable forms, then answer them in writing and stop. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Reflection differs from rumination in that reflection…** a. Ends in a decision or a change b. Is written down rather than thought c. Focuses on events rather than feelings d. Happens some time after the event **2. Eurich's research finds that most people…** a. Believe they are self-aware while not actually being so b. Are more self-critical than the evidence would justify c. Become more self-aware steadily as their career advances d. Can identify their own blind spots given enough time alone **3. Internal and external self-awareness differ in that external self-awareness is…** a. Knowing how others actually see you b. Awareness of your effect on a room c. Understanding your public reputation d. Knowing how you behave in a group **4. Asking "what did I contribute to this?" first matters because…** a. The answer is available to you and actionable b. Blaming others is generally considered unprofessional c. Your contribution is usually the largest single factor d. It makes the subsequent conversation less confrontational **5. A truth-teller is valuable specifically because they…** a. Say what they see, not what you want to hear b. Have known you for a long period of time c. Work closely enough to observe you daily d. Hold a more senior position than you do **6. Growth mindset, correctly stated, is the belief that…** a. Ability can be developed through effort and strategy b. Anyone can achieve anything if they want it enough c. Talent matters considerably less than hard work does d. Failure should always be reframed as a positive thing **7. Honest self-assessment matters MORE in the AI era because…** a. Knowing what you are actually good at is what lets you adapt b. Employers now assess self-awareness during hiring processes c. AI tools can measure your performance far more accurately d. Career changes have become considerably more frequent **8. Personality and strengths assessments are recommended here mainly as…** a. A prompt for reflection rather than a label b. A reliable measure of your own natural aptitude c. A shared vocabulary for describing a whole team d. A way of choosing which roles to apply for **9. A personal board of directors is…** a. A few people who hold you to your own goals b. A group of mentors drawn from your own industry c. The senior colleagues who assess all of your work d. A network you maintain for career purposes **10. The strongest evidence a values-behaviour gap audit was done honestly is…** a. It contains something you would rather not have written b. It covers all five of your stated values thoroughly c. It draws on evidence from the past three months d. It ends with a specific and achievable action ### Module 15 — Resilience: Recovering, Not Avoiding **Guiding question:** How do I recover from real setbacks fast enough to keep going — and treat that as a trainable skill? **Outcome:** Build recoverability: understand resilience as a trainable skill rather than a fixed trait, build a failure-response protocol before you need it, separate identity from results, maintain the physical baseline that psychological resilience rests on, and practise cognitive reframing to extract something usable from every setback. **AI-era lens:** Resilience is the most-cited quality in leadership research for sustained performance through public failure, and the AI era supplies a specific new stressor: the ground keeps moving. Skills you invested in genuinely do get automated; roles genuinely do change shape. Resilience is what lets you treat that as information to adapt to rather than as a verdict on your worth — which is why it sits in the character phase, not the technical one. **Apply-at-work mission — Failure-response protocol:** Write your protocol before you need it: when something goes badly wrong, who do you contact, what do you tell yourself, what do you do in the first 24 hours, and what is the physical baseline you protect? Then apply cognitive reframing to a real recent setback — find one genuinely useful learning from it, honestly, without pretending it was secretly good. **Reflection:** How do I actually behave in the first hours after a setback — and is that the behaviour I would advise someone else to have? What is my real recovery time? #### Resources - [Coursera — Resilience Skills in a Time of Uncertainty (UPenn)](https://www.coursera.org/learn/resilience-uncertain-times) — Karen Reivich, free to audit. Evidence-based resilience from the Penn Positive Psychology Center: thinking traps, real-time resilience, and the skills that hold up under pressure. - Identify your characteristic thinking traps under stress - Apply a real-time resilience technique in the moment - Build the relationships that resilience actually rests on - [Coursera — De-Mystifying Mindfulness (Leiden University)](https://www.coursera.org/learn/mindfulness) — Free to audit, and unusually rigorous — as much critical examination as practice. Useful precisely because it does not oversell the intervention. - Practise attention training without the mysticism - Understand what the evidence does and does not support - Use a short practice as a recovery tool, not a cure - [Viktor Frankl — Man's Search for Meaning](https://www.beacon.org/Mans-Search-for-Meaning-P602.aspx) — The advanced read: the account of surviving the camps and the argument that meaning, not comfort, is what carries people through. Short, and unlike anything else on this list. - Separate what happens from the stance you take toward it - Locate meaning that does not depend on outcomes - Recognise the limits of any resilience technique #### Project — Failure-response protocol + one real reframe (Portfolio component) Write your protocol before you need it. When something goes badly wrong: who do you contact in the first hour, what do you tell yourself, what do you deliberately not decide for 24 hours, and what is the physical baseline you protect no matter what? Make it specific enough to follow while upset. Then apply cognitive reframing to a real recent setback: find one genuinely useful learning from it, honestly, without pretending it was secretly a good thing. **Deliverable:** skills/m15-resilience-protocol.md — a written failure-response protocol plus one honest reframe of a real setback. Portfolio evidence for the character family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Followable while upset | 30% | Concrete named actions and named people, written so it works when your judgment is impaired. | | A 24-hour decision boundary | 20% | You name what you will NOT decide in the first day — the protection against a bad irreversible move. | | Physical baseline specified | 20% | Sleep, movement and food commitments stated concretely rather than as a general intention. | | A reframe that stays honest | 30% | One real learning from a real setback, without recasting the setback as secretly fortunate. | #### Scenario drills **Drill 1.** Something has gone badly wrong and you are composing an angry or defensive message. **Task:** Write it in full. Then ask an AI model how it will read to the recipient in a week, and what it will cost you. Do not send anything for 24 hours. **Drill 2.** You are stuck in catastrophising about something that has not happened yet. **Task:** Ask a model to help you write out the realistic worst case, the likely case, and your actual response to each. Keep the worst case concrete rather than vague. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Resilience is best understood as…** a. Recovering from difficulty, not avoiding it b. The ability to remain calm under any pressure c. A stable personality trait you are born with d. Preventing setbacks through careful planning **2. A failure-response protocol is written in advance because…** a. Judgment is worst exactly when you need it most b. Written plans are easier to share with a manager c. It reduces the chance of the failure happening d. Most organisations require an escalation process **3. Separating identity from results means…** a. A bad outcome is something that happened, not who you are b. Caring less about the outcomes of your work over time c. Keeping your professional and personal lives separate d. Refusing to accept personal blame for team failures **4. The physical baseline matters to resilience because…** a. Psychological recovery rests on it b. Exercise directly improves your mood c. It is the easiest thing to control d. Sleep loss is a well-known stressor **5. Cognitive reframing, done honestly, means…** a. Finding a real learning without recasting the event as good b. Choosing to focus on the positive aspects of what occurred c. Deciding the setback will not matter in five years' time d. Comparing your situation with people who have it worse **6. The AI-era stressor this module names is…** a. The ground keeps moving under invested skills b. Constant availability and blurred working hours c. Increased competition for entry-level positions d. Pressure to adopt tools faster than is comfortable **7. A thinking trap is…** a. A recurring distortion your mind runs under stress b. A decision made without sufficient information c. A belief held despite contradicting evidence d. An assumption you have never examined **8. Post-mortems that ask "what did we learn?" rather than "who failed?" tend to…** a. Surface the information the other version suppresses b. Take considerably less time to run than the alternative c. Make people feel better about a bad outcome d. Avoid the difficult conversations that need having **9. Frankl's central claim is that…** a. The stance you take toward suffering remains yours b. Suffering is meaningful and should not be avoided c. Optimism is what determines who survives hardship d. Meaning comes from the goals you eventually achieve **10. The most useful thing to know about your own recovery is…** a. How long it actually takes you b. Which setbacks affect you most c. Who you turn to when it happens d. What triggered the setback itself ### Module 16 — Patience: Playing the Long Game **Guiding question:** How do I resist the pull to rush — and make decisions on a decade horizon rather than a quarterly one? **Outcome:** Build strategic patience: recognise the rush urge and the comparison trap for what they are, shift your operating timeframe deliberately, distinguish healthy urgency about quality from unhealthy impatience about ego, and build patience into how you make and communicate decisions. **AI-era lens:** Behavioural research consistently finds deferred gratification predicts long-run outcomes more reliably than raw speed. The AI era makes this harder and more valuable at once: everything is faster, everyone appears further ahead, and the comparison surface is now infinite. Patience is what stops a genuinely good long-term strategy being abandoned in month three because a peer posted a win. **Apply-at-work mission — Ten-year decision reframe:** Write a one-page ten-year vision. Then take three decisions you are currently facing and re-decide each one asking "what would I choose if my horizon were ten years rather than ten months?" Note which decisions changed. Separately, name the specific comparison that most destabilises you, and write what you actually know and do not know about that person's situation. **Reflection:** Which of my current anxieties is genuinely about quality, and which is really about comparison? What would I do differently if nobody could see my progress? #### Resources - [Coursera — A Life of Happiness and Fulfilment (Indian School of Business)](https://www.coursera.org/learn/happiness) — Raj Raghunathan, free to audit. Strong specifically on the comparison trap and the pursuit of superiority — the two forces this module is about resisting. - Recognise the comparison trap as it operates on you - Distinguish healthy ambition from status-chasing - Choose a measure of progress that is not relative - [Walter Mischel — The Marshmallow Test](https://www.hachettebookgroup.com/titles/walter-mischel/the-marshmallow-test/9780316230865/) — From the researcher himself, including the caveats the popular version drops — the effect is real, smaller than the folklore, and the strategies are teachable. - State what the research does and does not establish - Use the distancing strategies that actually work - Design an environment that reduces the need for willpower - [Eric Jorgenson — The Almanack of Naval Ravikant](https://www.navalmanack.com/) — The advanced read, and free online. Compounding applied to reputation, knowledge and relationships — the clearest short statement of decade-level thinking available. - Identify what compounds in your own life - Recognise the games that cannot be won by playing faster - Set a horizon long enough to change your choices #### Project — Ten-year reframe + the comparison audit (Portfolio component) Write a one-page ten-year vision — specific enough to be wrong. Then take three decisions you are currently facing and re-decide each one asking what you would choose if your horizon were ten years rather than ten months. Note which decisions changed and which did not. Separately, name the specific comparison that most destabilises you — a person, a peer group, a feed — and write down what you actually know and do not know about that person's situation. **Deliverable:** skills/m16-long-game.md — a ten-year vision, three decisions re-decided on a decade horizon, and an honest comparison audit. Portfolio evidence for the character family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | A vision specific enough to be wrong | 25% | Concrete enough that you could later judge whether it happened, rather than a statement of general aspiration. | | Three real decisions re-decided | 30% | Live decisions genuinely re-examined on the longer horizon, with the ones that did not change also recorded. | | The comparison named specifically | 25% | The actual person or group, not a general observation about social media being unhelpful. | | Known versus assumed separated | 20% | What you actually know about their situation set against what you have been assuming. | #### Scenario drills **Drill 1.** You are about to make a decision under a sense of urgency you cannot fully justify. **Task:** Ask an AI model to interview you about the decision on three horizons — ten weeks, ten months, ten years — and to state where the answers diverge. **Drill 2.** You feel behind relative to someone whose progress you keep watching. **Task:** Write down everything you actually know about their situation, then ask a model to list what a full account would also need to include. Notice how much is missing. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Strategic patience means…** a. Choosing on a longer horizon deliberately b. Waiting until you feel certain about it c. Taking more time over every decision d. Avoiding decisions that cannot be undone **2. The comparison trap is described as destructive mainly because…** a. The comparison surface is now effectively infinite b. People generally compare themselves with the wrong peers c. Competitive feelings interfere with team collaboration d. It causes people to overestimate their own progress **3. Healthy urgency differs from unhealthy impatience in that healthy urgency is about…** a. Quality, where impatience is about ego b. Deadlines that have real consequences c. Work you find genuinely interesting d. Things within your own control today **4. The marshmallow research, stated accurately, shows that…** a. Delay strategies are learnable and the effect is modest b. Willpower in childhood determines later life outcomes c. Children who wait are more intelligent than those who do not d. Deferred gratification matters more than any other trait **5. Patience is called physiological before psychological because…** a. The rush urge is a bodily state you can regulate b. Physical exercise makes people generally calmer c. Breathing techniques work faster than reasoning d. Tiredness is the main cause of impatient decisions **6. The ten-year vision is revisited quarterly in order to…** a. Calibrate the short-term choices against it b. Check whether the vision is still accurate c. Measure the progress you have made so far d. Keep motivation high during difficult periods **7. Something that compounds is best described as…** a. A gain that produces further gains on itself b. An investment that appreciates in value slowly c. A skill that improves with regular practice done d. A benefit that arrives long after the effort is made **8. The AI era makes patience harder specifically because…** a. Everything is faster and everyone appears further ahead b. Career paths have become substantially less predictable now c. Employers now expect their results on much shorter timescales d. The pace of technological change keeps accelerating **9. A good test of whether an anxiety is about quality or comparison is…** a. Whether you would still feel it unobserved b. Whether the work meets your own standard c. Whether the deadline is genuinely fixed d. Whether the anxiety is proportionate **10. Celebrating deliberate decisions as well as fast ones matters because…** a. Teams copy whatever behaviour actually gets praised b. Deliberate decisions are usually the better decisions c. It slows a culture that has become too reactive d. Speed is easier to measure than judgment is ### Module 17 — Integrity: Values Under Pressure **Guiding question:** Do I do what I said I would when it is costly — and have I even written down what I stand for? **Outcome:** Build values-alignment: document your core values explicitly, test integrity in low-stakes conditions, practise declining opportunities that conflict with your values, build a personal decision framework of things you will not do regardless of incentive, and make your reasoning transparent when decisions are hard. **AI-era lens:** Values-alignment under pressure is the foundation of trust and durable influence — and the AI era raises the stakes in a very concrete way. It is now trivially easy to generate work you did not do, claim understanding you do not have, and produce persuasive content you do not believe. Every one of those is an integrity decision that did not exist at this scale before, and each is made privately, which is exactly where integrity is actually tested. **Apply-at-work mission — Your "I will not" list:** Write your core values explicitly, then convert them into a decision framework: specific things you will not do regardless of the incentive. Include at least two that relate directly to AI use — what you will not pass off as your own, what you will not claim to understand. Then find a real recent decision where you were tested, and assess yourself honestly. **Reflection:** Where have I already quietly crossed a line I would not defend publicly? What would it cost me to hold the line next time, and am I willing to pay it? #### Resources - [Coursera — Ethical Leadership Through Giving Voice to Values (UVA Darden)](https://www.coursera.org/learn/uva-darden-giving-voice-to-values) — Mary Gentile, free to audit. The premise is unusual and useful: assume you already know what is right, and practise how to actually say it under pressure. - Rehearse the script for a values conflict in advance - Anticipate the standard rationalisations and answer them - Act on a value when acting is costly - [Michael Sandel — Justice (Harvard, free online)](https://justice.harvard.edu/) — The full lecture series, free. Not a leadership course — a serious moral philosophy one, which is precisely why it sharpens your ability to say why you hold a position. - Locate your own reasoning in the major moral traditions - Argue a values position without appealing to feeling - Recognise where your principles actually conflict - [Ray Dalio — Principles: Life and Work](https://www.principles.com/) — The advanced read on writing your decision rules down and holding to them. Read critically — the value is the practice of codifying principles, not adopting his. - Write decision principles you can apply under pressure - Distinguish a principle from a preference - Review and revise principles honestly over time #### Project — Your "I will not" list (Portfolio component) Write your core values explicitly, then convert them into a decision framework: specific things you will not do regardless of the incentive offered. Include at least two that relate directly to AI use — what you will not pass off as your own work, and what you will not claim to understand. Then find one real recent decision where you were actually tested, and assess yourself honestly against your own list, including if you fell short. **Deliverable:** skills/m17-integrity-framework.md — stated values, an "I will not" list including two AI-specific commitments, and an honest self-assessment against one real tested decision. Portfolio evidence for the character family. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Commitments are specific and costly | 30% | Each line names something you might plausibly be offered, not a refusal to do things nobody would ask. | | Two genuine AI commitments | 25% | Concrete lines about attribution and claimed understanding, applicable to the work you actually do. | | One real tested decision | 25% | An actual situation where the value cost you something, or where you discovered it did not hold. | | Assessed honestly | 20% | Where you fell short is recorded as falling short, without the surrounding justification. | #### Scenario drills **Drill 1.** You are drafting your "I will not" list and it feels comfortable. **Task:** Ask an AI model to generate the specific offers or pressures that would make each line genuinely hard for someone in your role. Rewrite any line that no plausible pressure threatens. **Drill 2.** You are about to use AI-produced work in something you will put your name to. **Task:** Before submitting, ask yourself — in writing — which parts you could defend under questioning. Use a model to interrogate you on the parts you claim to understand. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Integrity, as this module defines it, is…** a. Acting on your values when it is costly b. Being honest in your dealings with others c. Keeping your promises to other people d. Behaving consistently across all situations **2. Documenting your values explicitly matters because…** a. Undocumented values bend under pressure b. It is a common requirement for leadership roles c. Written statements are easier to share with a team d. Most people cannot recall their values when asked **3. The "Giving Voice to Values" approach assumes that…** a. You already know what is right and need to practise saying it b. Most ethical failures come from not knowing the right answer c. Ethical reasoning improves with formal philosophical training d. People behave better when the consequences are made clear **4. Testing integrity in low-stakes conditions is useful because…** a. The habit is what carries you through high-stakes ones b. Low-stakes tests are more objective and easier to score c. It builds a documented record of consistent behaviour d. Small compromises are more common than large ones **5. An "I will not" list is stronger than a values statement because…** a. It converts a value into a decidable rule b. It is shorter and therefore easier to recall c. Negative commitments are harder to rationalise d. It states clearly what you stand against **6. The AI-specific integrity decisions this module names are about…** a. Attribution and claimed understanding b. Data privacy and confidentiality c. Tool selection and vendor lock-in d. Bias and fairness in model output **7. These integrity decisions are described as newly consequential because…** a. They are made privately, where integrity is actually tested b. The technology involved is difficult for most people to grasp c. Organisations have not yet written policies covering them d. The penalties for getting them wrong have grown severe **8. Making your reasoning transparent on hard decisions mainly helps by…** a. Showing the values calculus rather than only the outcome b. Protecting you if the decision later turns out to be wrong c. Demonstrating that you considered every available option d. Allowing other people to contribute before you decide **9. Punishing admitted mistakes damages integrity culture because…** a. It makes concealment the rational choice b. It discourages people from taking any risks c. It creates resentment towards the leadership d. It slows down organisational decision-making **10. A personal decision framework differs from a preference in that it…** a. Holds when the incentive runs the other way b. Is written down in a formal document c. Applies across your work and personal life d. Has been discussed with people you trust ### Module 18 — Public Service Drive: Purpose Beyond Yourself **Guiding question:** What am I actually for — and where do my skills meet a need bigger than my own advancement? **Outcome:** Build durable purpose: clarify the why beneath your work, distinguish service motivation from self-sacrifice, find the overlap between your skills and real needs, practise leadership in a community or civic context, and apply service-first thinking to commercial work. **AI-era lens:** Research in political science finds leaders motivated by duty and service make more consistent long-term decisions and earn more durable trust than those driven by personal ambition alone. This closes the character phase deliberately: in an era where capability is increasingly abundant and cheap, the differentiating question stops being "what can you do?" and becomes "what are you doing it for?" Purpose is also the most reliable fuel for the persistence the other sixteen skills demand. 🎯 Completes Capstone Milestone 5. **Apply-at-work mission — Skills-meets-need map:** Map your genuine skills against needs you actually care about, and find the overlap — the place where what you are good at meets something that matters beyond your own advancement. Commit to one concrete action within it this month: a contribution, a role, a project. Then write the version of your work story that leads with purpose rather than position. **Reflection:** Would I still do this work if it earned me no status? What need do I keep noticing that nobody seems to be addressing? #### Resources - [Coursera — Social Entrepreneurship (Copenhagen Business School)](https://www.coursera.org/learn/social-entrepreneurship) — Free to audit. How service-driven ventures are actually built and sustained — the practical counterweight to purpose as a feeling. - Frame a real need as something addressable - Test whether your contribution is actually wanted - Design service work that is durable, not heroic - [Coursera — Social Impact Strategy (University of Pennsylvania)](https://www.coursera.org/learn/social-impact) — Free to audit. Strategy and design thinking applied to social problems, including how to tell real impact from activity that feels meaningful. - Distinguish measured impact from felt impact - Build a theory of change for a real initiative - Choose where your effort has the most leverage - [Simon Sinek — Start With Why](https://simonsinek.com/books/start-with-why/) — The advanced read on purpose as an organising principle. Read it critically — the framework is more useful than the case studies, several of which have aged awkwardly. - Articulate your own why in a defensible sentence - Distinguish purpose from positioning - Lead with purpose without overclaiming it #### Project — Skills-meets-need map + one commitment (Portfolio component) Map your genuine skills — the ones you would defend, not the ones you wish you had — against needs you actually care about. Find the overlap: where what you are good at meets something that matters beyond your own advancement. Commit to one concrete action within it this month: a contribution, a role, a project, with a date. Then write the version of your work story that leads with purpose rather than position, and check whether it is still true when read by someone who knows you. **Deliverable:** skills/m18-purpose-map.md — a skills-and-needs map, one dated commitment, and a purpose-led version of your work story. Portfolio evidence for the character family. 🎯 Completes Capstone Milestone 5. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Skills stated honestly | 25% | Capabilities you could demonstrate on request, not aspirations described in the present tense. | | Needs that are genuinely yours | 20% | Things you already notice and return to, rather than causes chosen because they sound worthy. | | One dated, concrete commitment | 30% | A specific action with a date and a first step — not an intention to get involved at some point. | | A purpose-led story that survives a reader | 25% | Written so someone who knows you would recognise it as true rather than as positioning. | #### Scenario drills **Drill 1.** You want to find where your skills and something you care about actually overlap. **Task:** List your ten strongest capabilities and the five problems you keep noticing. Ask an AI model to propose overlaps you have not considered, then discard the ones you do not actually want. **Drill 2.** You have written a purpose statement and you are not sure it is honest. **Task:** Ask a model to read it as a sceptical colleague who has worked with you for three years, and to name which claims your behaviour would need to support. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Service motivation differs from self-sacrifice in that service motivation is…** a. More durable, because it is not paid for in depletion b. Much more visible to all of the people around you at work c. Directed at institutions rather than individuals d. Sustained by recognition from the community **2. The research finding this module rests on is that duty-driven leaders…** a. Make more consistent long-term decisions b. Are promoted more often than their peers c. Report higher satisfaction with their work d. Build larger organisations over their careers **3. The skills-meets-need overlap is described as your highest-leverage contribution because…** a. Capability without care fades and care without capability frustrates b. It is where the most funding and support are usually available c. Working on things you enjoy makes you considerably more productive d. Employers increasingly look for purpose-aligned candidates **4. Purpose closes the character phase because…** a. It is the most reliable fuel for the persistence the other skills need b. It is the hardest of the character traits to develop properly c. The other traits are prerequisites for finding your purpose d. It is the trait employers most frequently ask about **5. The question this module says replaces "what can you do?" is…** a. What are you doing it for? b. Who benefits from your work? c. What would you do for free? d. Where do you want to end up? **6. A theory of change is…** a. An account of how your action leads to the outcome b. A plan for introducing change into an organisation c. A framework for measuring social impact reliably d. A statement of the values driving an initiative **7. Community or civic leadership is recommended as a practice ground because…** a. It offers real stakes without career risk b. It looks strong on a professional CV afterwards c. It is easier than leading at work is d. It provides access to a wider network **8. Leading with purpose rather than position means describing your work by…** a. What it is for, rather than what your title is b. The impact numbers rather than the actual activities c. Your values rather than your listed achievements d. The team rather than your own contribution **9. The main risk in a purpose-led work story is that it…** a. Becomes positioning rather than description b. Sounds rather naive to a commercial audience c. Sets expectations that you cannot meet later d. Understates your actual accomplishments **10. Service-first thinking applied commercially claims that organisations which serve…** a. Tend to outlast those that extract b. Are more profitable in the short run c. Attract better staff than competitors d. Face less regulatory pressure overall --- ## Phase 5 — Integrate (weeks 10) ### Module 19 — Capstone: Your Skills Portfolio & 12-Month Plan **Guiding question:** What have I actually changed — and what is my deliberate plan for the next twelve months? **Outcome:** Assemble everything into a portfolio: re-take the baseline self-assessment and compare honestly, gather the evidence artifacts produced across all modules, take one skill to demonstrated depth, and build a specific 12-month development plan with quarterly focus areas, chosen resources, and a review rhythm you will actually keep. **AI-era lens:** A course like this fails if it becomes seventeen interesting ideas you agreed with and forgot. The portfolio is the mechanism that prevents that: measured before-and-after on every skill, real artifacts as evidence rather than claims, one skill proven at depth, and a dated plan with review points. What you leave with is not a certificate — it is a document that makes your development deliberate instead of accidental. **Apply-at-work mission — Portfolio + defended plan:** Re-take the Module-1 baseline across all 17 skills and place the two side by side, honestly — including anything that did not move. Assemble your artifacts as evidence. Choose one skill and take it to demonstrated depth with real proof. Then write a 12-month plan: quarterly focus, specific resources, practice habits, and how you will review it. Share it with one person who will hold you to it. **Reflection:** Reopen the Module-1 sealed baseline: where was I wrong about myself on day one? Which skill turned out to matter more than I expected, and which less? #### Resources - [WEF — Future of Jobs Report 2025](https://www.weforum.org/publications/the-future-of-jobs-report-2025/) — Re-read the skills-outlook chapter now, having done the course. The same table reads differently once you have a measured baseline and nine weeks of practice behind you. - Re-locate your own skills on the report's outlook - Identify which rising skills you have now touched - Choose next-year priorities against external evidence - [Peter Drucker — Managing Oneself (Harvard Business Review)](https://hbr.org/2005/01/managing-oneself) — The classic short essay on knowing your strengths, how you work, and what you contribute — and the feedback-analysis method that produces those answers rather than guessing them. - Apply feedback analysis to a decision you have made - State how you actually work best, with evidence - Define the contribution you intend to make next - [Coursera — Learning How to Learn](https://www.coursera.org/learn/learning-how-to-learn) — Returning here deliberately. Your twelve-month plan is a learning system, and the mechanics from Module 3 are what will decide whether it survives March. - Design a review rhythm you will actually keep - Space the plan's practice rather than front-loading it - Build retrieval into your development habits #### Project — Skills Portfolio + defended 12-month plan (Capstone) Four parts. First, reopen the sealed Module-1 baseline, re-take it across all 17 skills, and place the two side by side honestly — including everything that did not move and anything you now rate lower because you understand the skill better. Second, assemble your module artifacts as evidence, with a one-line note on what each one demonstrates. Third, take one skill to demonstrated depth: real proof that someone else could examine, not a claim. Fourth, write a 12-month development plan — quarterly focus areas, named resources, practice habits, and dated review points — and share it with one person who will actually hold you to it. **Deliverable:** skills/m19-portfolio.md — baseline versus final across 17 skills, the artifact index, one skill demonstrated at depth, and a dated 12-month plan with an accountability partner named. The completed Skills Portfolio. **Assessment rubric** | Criterion | Weight | What good looks like | | --- | ---: | --- | | Honest before-and-after across all 17 | 25% | Both assessments side by side, including the skills that did not move and any rating that went down on better understanding. | | Artifacts assembled as evidence | 20% | The module deliverables indexed, each with a line on what it actually demonstrates rather than what it was titled. | | One skill demonstrated at depth | 25% | Proof someone else could examine — a taught session, a shipped piece of work, a documented outcome — not a self-report of fluency. | | A plan with quarters, resources and review dates | 20% | Specific enough to follow: what, when, using which resource, reviewed on which dates. | | Shared with an accountability partner | 10% | A named person who has seen the plan and agreed to ask about it — the mechanism that makes the dates real. | #### Scenario drills **Drill 1.** You have both assessments in front of you and the comparison feels flattering. **Task:** Ask an AI model to interrogate three of your improved ratings: what evidence supports each, and what would someone who works with you say? Downgrade any rating you cannot defend. **Drill 2.** Your twelve-month plan is written and it looks reasonable. **Task:** Ask a model to predict which quarter it fails in and why, given how you have described your actual life. Then fix the plan rather than the prediction. #### Knowledge check (10 questions) **Self-test prompts. Answers and explanations are not published here — take the quiz at https://ragentic.netlify.app/#/courses/essential-skills to check yourself.** **1. Re-taking the baseline matters mainly because…** a. Felt progress and measured progress often disagree b. A second assessment is more accurate than the first c. It provides a score to put on your final certificate d. Repetition helps you remember the seventeen skills **2. A rating that goes DOWN between baseline and final usually means…** a. You now understand the skill well enough to judge it b. The skill genuinely deteriorated during the course c. The original self-assessment was carelessly completed d. You have become unnecessarily harsh on yourself **3. Artifacts are required as portfolio evidence because…** a. A claim about a skill is not evidence of it b. They demonstrate that the work was completed c. They are easier to assess than self-ratings are d. Employers increasingly ask to see work samples **4. "Demonstrated depth" in one skill means…** a. Proof another person could examine b. A rating of fluent on that skill c. More practice hours than the others d. A completed course with a certificate **5. The plan is built around quarterly focus areas rather than all 17 skills because…** a. Focused improvement beats diffuse effort b. Four is the standard planning horizon c. Some of the seventeen matter far less d. Reviewing quarterly is easier to sustain **6. Dated review points matter because…** a. A plan with no review date quietly stops existing b. Regular reviews keep motivation levels high c. They create a record of what you attempted d. Most development plans require formal sign-off **7. Sharing the plan with one person mainly works because…** a. Someone will ask about it when you would not b. Explaining it forces you to clarify the whole plan c. Other people can suggest some useful improvements d. It creates a social cost to giving up on it **8. The course argues its own failure mode would be…** a. Seventeen interesting ideas you agreed with and forgot b. Learners finding the material too demanding to finish c. Too much focus on human skills over technical ones d. Coverage of skills that are hard to apply at work **9. The AI-era thesis, restated at the end, is that value shifts toward…** a. Deciding, judging, and moving people to act b. Technical fluency with the newest available tools c. Speed of production across every kind of work d. Specialised knowledge in a defensible domain **10. The most valuable thing you leave the course with is…** a. A document that makes your development deliberate b. A certificate demonstrating that you completed it c. A vocabulary for describing seventeen useful skills d. A set of resources you can return to when needed ## Toolkits ### 17-skill self-assessment matrix **Unlocks in module 1.** The baseline you take in Module 1, seal, and reopen in Module 19. Four levels, an evidence sentence for every rating, and space for the second pass beside the first. ```markdown # 17-skill self-assessment matrix Rate each skill on the four-level scale. **Every rating needs one sentence of real evidence** — a specific situation, not an impression. A rating without evidence is a feeling about yourself. **The scale** | Level | Means | |---|---| | 1 · Novice | I know the term. I have not deliberately practised it. | | 2 · Practising | I use it sometimes, unevenly, usually when I remember to. | | 3 · Fluent | I use it reliably under normal conditions, including some pressure. | | 4 · Could teach it | I could bring another person to genuine competence in it. | Level 4 is deliberately hard. Ask yourself whether you could get someone else to level 3 — not whether you feel confident. ## The matrix | # | Skill | Family | Baseline (M1) | Evidence — one real situation | Final (M19) | Δ | |---|---|---|---|---|---|---| | 1 | Metacognition | Cognitive | | | | | | 2 | Learnability | Cognitive | | | | | | 3 | Critical thinking | Cognitive | | | | | | 4 | Financial literacy | Cognitive | | | | | | 5 | Storytelling | Cognitive | | | | | | 6 | Prompting & vibe coding | AI fluency | | | | | | 7 | AI workflow integration | AI fluency | | | | | | 8 | Human behaviour | Interpersonal | | | | | | 9 | Empathy & compassion | Interpersonal | | | | | | 10 | Feedback fluency | Interpersonal | | | | | | 11 | Emotional vulnerability | Interpersonal | | | | | | 12 | Conflict resolution | Interpersonal | | | | | | 13 | Self-reflection | Character | | | | | | 14 | Resilience | Character | | | | | | 15 | Patience | Character | | | | | | 16 | Integrity | Character | | | | | | 17 | Public service drive | Character | | | | | ## Then, in Module 1 only - **Three priority skills** — where genuine improvement would most change your next twelve months, and *why those three*. Not the three that sound most impressive. - **Date it and seal it.** Do not look at it again until Module 19. The comparison is only worth anything if you cannot reconstruct what you originally wrote. ## Then, in Module 19 only - Fill the final column *before* rereading your baseline evidence. - **A rating that goes DOWN is a good sign**, not a bad one: it usually means you now understand the skill well enough to judge it properly. Say so explicitly in the write-up. - Record the skills that did not move. Those are the honest part of the portfolio. ``` ### Personal learning-system template **Unlocks in module 2.** Capture, spaced repetition, review rhythm and a rule for what deserves capturing at all — designed so you will still be running it in March. ```markdown # My personal learning system The system you will actually keep beats the ideal one you abandon in a fortnight. Elaborate systems collapse under their own overhead. Design for your worst week, not your best one. ## 1 · Capture | Question | My answer | |---|---| | Tool (Obsidian, Notion, paper, notes app) | | | Where it lives, so it is always to hand | | | What gets captured — the rule | | | What does NOT get captured | | **On the rule:** "interesting" is not a rule. Something like "only things I expect to need again within a year, in my own words" is. The second half matters — copied text is not captured. ## 2 · Retrieval (spaced repetition) | Question | My answer | |---|---| | Tool (Anki, RemNote, paper cards) | | | When I review — the trigger, not the intention | | | How many new items per day, realistically | | | What deserves memorising at all | | Retrieval is effortful by design. If reviewing feels comfortable, you are re-reading rather than recalling, and it is not working. ## 3 · Review rhythm | Cadence | What I do | When | |---|---|---| | Weekly | Process captures; delete what turned out not to matter | | | Monthly | Reread one older note; check it still makes sense | | | Quarterly | Prune. A system that only grows will be abandoned | | ## 4 · The teaching test The real check on understanding: **explain it to a person within 48 hours.** | Concept learned | Taught to whom | Date | What their questions exposed | |---|---|---|---| | | | | | The last column is the valuable one. Where their question stopped you is exactly where your understanding was borrowed rather than owned. ## 5 · The honest review, one month in - What part of this system have I silently stopped doing? - Was it too elaborate, badly triggered, or genuinely not useful? - What is the smaller version I would still be running in six months? ``` ### AI-practice prompt pack **Unlocks in module 4.** One sparring prompt per skill — seventeen ways to use a model as a practice partner rather than as a substitute. The through-line of the whole course, in one page. ```markdown # AI-practice prompt pack — one sparring prompt per skill The rule that governs all of these: **AI is the sparring partner, never the substitute.** Use it to rehearse, critique, generate options and interrogate you. Do not use it to have the conversation, hold the value, or feel the thing on your behalf. ## Cognitive **Metacognition** — before a decision you have already made up your mind about: > Argue the strongest case AGAINST my preferred option. Then list the points I would struggle > to answer. **Learnability** — after reading something complex: > I am going to explain this from memory, without the source. Grade my explanation against the > actual concept and list precisely what I missed. **Critical thinking** — on any answer you are about to rely on: > List every assumption this answer depends on. Then tell me which single assumption, if wrong, > would most damage the conclusion. **Financial literacy** — on a statement or a rule you half-understand: > Explain this to a smart beginner. Then tell me the question an expert would ask about it that > I have not thought to ask. **Storytelling** — on something true, useful and flat: > Where is the tension in this, and where would a reader stop caring? Do not rewrite it — just > tell me where it dies. ## AI fluency **Prompting & vibe coding** — when you know the goal but not the specification: > Before you answer, ask me the questions you need answered to do this well. **AI workflow integration** — on a recurring task: > Split these steps into three groups: safe to delegate, needs my review, must stay human. Then > defend the boundary you drew, because I am going to argue with it. ## Interpersonal **Human behaviour** — before concluding someone is being difficult: > Here is only what they did, with no adjectives. Give me five explanations, ranked by > likelihood, including structural ones. **Empathy & compassion** — before a conversation you are dreading: > Role-play this person, briefed with their situation and constraints — not my judgment of them. > Afterwards, tell me what I missed. **Feedback fluency** — on feedback you have drafted: > Find every place I have slipped a judgment in disguised as an observation. Leave only > observable behaviour. **Emotional vulnerability** — when unsure whether to share something: > Argue both readings: first that saying this serves the relationship, then that it burdens the > other person. Do not give me a verdict. **Conflict resolution** — when stuck on a single number or decision: > Here is everything both sides care about. Propose five trades across those issues that neither > of us has raised. ## Character **Self-reflection** — on a self-assessment that feels flattering: > Write the version a well-informed critic who dislikes me would write from these same facts. **Resilience** — when catastrophising: > Help me write the realistic worst case, the likely case, and my actual response to each. Keep > the worst case concrete, not vague. **Patience** — on a decision made under unexplained urgency: > Interview me about this decision on three horizons — ten weeks, ten months, ten years — then > tell me where my answers diverge. **Integrity** — on your "I will not" list: > For each line, invent the specific offer or pressure that would make it genuinely hard for > someone in my role. **Public service drive** — on a purpose statement: > Read this as a sceptical colleague who has worked with me for three years. Which claims would > my actual behaviour need to support? ## The discipline After each of these, write your own answer *before* accepting the model's framing. The prompt exists to produce something for you to react to — not to produce the conclusion. ``` ### 12-month development plan **Unlocks in module 10.** Quarterly focus, named resources, practice habits with triggers, dated review points, and one person who has agreed to ask. The Module-19 deliverable that decides whether any of this survives contact with next year. ```markdown # My 12-month development plan Written in Module 19, from the baseline-versus-final comparison. Coverage was the course's job. Focus is now yours: two or three skills a quarter, not seventeen. ## The starting position - **Skills that moved most:** - **Skills that did not move:** - **Anything I now rate LOWER because I understand it better:** - **The one skill I took to demonstrated depth, and the proof:** ## Quarterly focus | Quarter | Focus skills (2–3) | Why these, now | Named resource | Practice habit + its trigger | |---|---|---|---|---| | Q1 | | | | | | Q2 | | | | | | Q3 | | | | | | Q4 | | | | | **On the habit column:** an intention is not a habit. "Review my decision journal on Friday afternoon before I shut the laptop" is a habit, because it names when. A habit with no trigger is a hope. **On the resource column:** name the actual course, book or practice — not "read more about it". Anything you were not going to finish should not be in the plan. ## Review points | Date | What I check | Done? | |---|---|---| | | Did the quarter's habit actually run? | | | | Has anything changed that makes this plan wrong? | | | | Re-rate the quarter's focus skills, with evidence | | | | Full 17-skill re-assessment, one year on | | Put these in your calendar now, with a reminder. The failure mode is not abandoning the plan deliberately — it is never returning to it and never noticing that you did not. ## Accountability - **Person who has agreed to ask me about this:** - **What I have asked them to ask:** - **When:** One specific question arriving at a moment when you had stopped thinking about it is worth more than any amount of resolve. ## The honest prediction Which quarter does this plan fail in, and why? Write it down now — the over-full quarter, the resource you will not finish, the habit with no real trigger. Then fix the plan rather than the prediction. ``` ## Capstone ### Skills Portfolio — depth choice · Take a cognitive skill to depth Metacognition, learnability, critical thinking, financial literacy or storytelling, proven rather than claimed: a decision journal with closed prediction loops, a taught concept with the learner's questions recorded, a published steelman, a read annual report, or a story that someone else retold. ### Skills Portfolio — depth choice · Take an AI-fluency skill to depth Prompting or workflow integration, evidenced: a prompt library in real use with documented failure modes, or one workflow rebuilt end to end with before-and-after numbers, an evaluation rubric written in advance, and an honest account of what the change cost. ### Skills Portfolio — depth choice · Take an interpersonal skill to depth Human behaviour, empathy, feedback, vulnerability or conflict — the hardest family to evidence and the most valuable when you can. A feedback exchange that changed something, a resolved disagreement with the trade named, or a motivation map that proved out. ### Skills Portfolio — depth choice · Take a character trait to depth Self-reflection, resilience, patience, integrity or purpose. Evidence here is behavioural over time: a values-behaviour audit checked by a truth-teller, a failure protocol that was actually used, or a commitment you kept when it cost you something. ### Skills Portfolio — depth choice · Teach one skill to someone else The highest bar on the four-level scale, available for any of the seventeen: design and run a real session that brings another person to genuine competence. Their questions and what they could do afterwards are the evidence. ### Your own · Your own demonstration The strongest depth proof is one your actual life demanded. If a real situation this term forced you to use one of these skills properly, document that — the stakes were genuine, which no designed exercise can match. ### Milestones - **Module 1 — Baseline self-assessment** - **Module 2 — Cognitive family evidence** - **Module 4 — AI fluency evidence** - **Module 7 — Interpersonal family evidence** - **Module 9 — Character family evidence** ### Portfolio checklist - M1 — the dated, sealed baseline across all 17 skills, with evidence sentences - M2 — decision journal: five real decisions with reasoning, confidence and falsifiers, plus bias tripwires - M3 — your personal learning system, run on something real and taught within 48 hours - M4 — a steelman of your own position, and an audited AI answer with assumptions surfaced - M5 — one annual report read and summarised in plain language, plus a net-worth tracking habit - M6 — a story bank of ten tagged anecdotes, and one dry piece rewritten to earn attention - M7 — a prompt library of eight or more, each with its pattern, an example and its failure mode - M8 — one workflow rebuilt with AI, measured before and after, costs included - M9 — a week of observation before diagnosis, and a motivation map with a stated-versus-revealed gap - M10 — three listening-only conversations, and an empathy map written in the other person's voice - M11 — one hard piece of feedback given, one specifically requested, and your own reaction recorded - M12 — something real shared with someone whose trust matters, and your own boundary written - M13 — a live disagreement mapped to interests, with a trade positions would never have found - M14 — a values-behaviour gap audit, checked against a truth-teller - M15 — a failure-response protocol, plus one honest reframe of a real setback - M16 — a ten-year vision, three decisions re-decided on that horizon, and a comparison audit - M17 — your "I will not" list, including two AI commitments, tested against a real decision - M18 — a skills-meets-need map with one dated commitment, and a purpose-led work story - M19 — baseline versus final across all 17, one skill demonstrated at depth, and a dated 12-month plan shared with someone who will ask about it