Teach on Udemy

Turn what you know into an opportunity and reach millions around the world.

Learn More

Your cart is empty.

Keep shopping

Critical Thinking with AI

Build the AI judgment skills that make your decisions accurate, defensible, and trusted, before a costly mistake define

Role Play

New

Created byVani Aggarwal, Agile Enterprise Coach London

Last updated 6/2026

English

What you'll learn

Identify the psychological patterns that cause professionals to accept flawed AI outputs — and apply a structured audit system to every significant AI-assisted
Apply a professional-grade decision framework to determine exactly how much trust any AI recommendation deserves — based on stakes, data recency, domain, and re
Evaluate AI-generated strategies and recommendations using advanced reasoning techniques that expose hidden assumptions, map downstream consequences, and detect
Design your own AI supervision practice — including documentation habits, team protocols, and a weekly review system — that compounds your judgment and professi

Course content

8 sections • 32 lectures • 50m total length

Welcome: Two Types of Professional, One Course1:59
There are two distinct types of professional emerging in the AI era — and only one of them is building a career advantage that compounds. This lecture names the difference, tells you exactly which type this course builds, and sets the stakes for everything that follows. If you have ever wondered whether your AI use is actually making you better at your job, this is where that question gets answered.
The Hidden Skill No One Is Teaching1:33
Every AI course teaches you to use the tools. Almost none of them teach you what to do with what comes back. This lecture identifies the professional skill that sits between an AI output and a sound decision — a skill most organisations are deploying AI without. Understanding it changes how you see every AI interaction you will have from this point forward.
Your Transformation: From AI User to AI Supervisor0:49
There is a specific professional identity emerging in AI-augmented organisations — and it is not the person who uses AI most. This lecture defines what that identity looks like, what capabilities it requires, and what it means for your career trajectory. The distance between where you are and where you need to be is shorter than you think.

The Great Skill Shift: From Doing to Supervising1:27
Something fundamental has changed about what organisations need from their best people — and it is not what most professionals expect. This lecture maps the shift that is redefining career value in every industry, and names the specific capability gap most professionals do not know they have. The professionals who close this gap first are the ones accumulating the most durable career advantage right now.
The Automation Bias Trap2:20
There is a psychological mechanism in every human brain that AI triggers more powerfully than almost any tool before it — and it is responsible for some of the most expensive professional errors of the past five years. This lecture traces that mechanism through a real organisational case that played out over four years before anyone caught it. What you discover here will permanently change how you feel when an AI output lands in your inbox.
The Confidence-Competence Inversion2:10
In humans, confidence is usually a reliable signal of competence. In AI, that relationship breaks down completely — and the consequences for professionals who do not know this are significant. This lecture names the phenomenon, shows it in action through a real professional scenario, and introduces the core habit that distinguishes AI supervisors from AI users.
Recognising the Gap

Why 90% of AI Failures Happen Before the Prompt1:04
The instinct when an AI output disappoints is to blame the prompt. The reality is almost always more upstream than that — and much easier to fix. This lecture identifies the actual origin point of most AI output failures and introduces the professional discipline that eliminates them at the source. The shift is smaller than you expect and the impact on output quality is immediate.
Defining the Success North Star1:17
AI produces what you ask for. The problem is that most professionals have not fully defined what they actually need before they ask. This lecture introduces a pre-prompt practice used by the professionals who consistently get the most accurate, useful AI outputs — and explains why skipping it is the single most common cause of wasted AI interactions in professional settings.
Context Injection: Feeding Tribal Knowledge to AI1:06
AI is trained on the world's information. It knows nothing about your organisation, your client, your team, or the unwritten rules that govern your specific situation. This lecture shows how to close that gap — and why the professionals who do this consistently get outputs that require dramatically less revision, correction, and professional embarrassment.
Workshop: The Prompt Makeover0:46
This is where the theory of the last three lectures becomes a skill you can use before you close this window. A real professional request gets transformed step by step — and the difference in output quality is visible within minutes. This lecture is the one most participants cite as the fastest practical return in the course.

Introducing VERIFY: Your Six-Step Audit System2:18
This is the framework at the centre of this course — and the one that changes how you interact with AI outputs from the moment you finish watching. Six steps. Each one targets a specific failure mode. Together they form a complete quality system that can be run in under ten minutes on any significant AI-assisted output. This lecture introduces the full architecture before each step is built out in detail.
The Seduction of Fluency1:09
AI writes the way experts write — structured, confident, authoritative. That is not an accident. And it is not necessarily a signal that the content is correct. This lecture exposes a specific psychological response that professional-quality AI writing triggers in the human brain — one that is responsible for more uncaught errors than any other single factor in AI-assisted professional work.
Hallucination Hunting: Three Technical Tells1:12
AI hallucinations do not look like mistakes. They look like facts. This lecture identifies three specific patterns that appear consistently in AI-generated inaccuracies — patterns that, once you know them, are detectable in under two minutes. These are the three checks that professional AI supervisors run before anything significant leaves their desk.
Bias Auditing: Identifying the Corporate Default1:03
AI outputs are not neutral. They reflect the dominant patterns in their training data — and in professional contexts, those patterns carry a specific set of assumptions that may or may not apply to your situation. This lecture names the bias pattern most likely to affect your work, shows how to recognise it, and gives you the audit question that catches it before it reaches your client or your leadership team.
Reverse-Test and Find the Counterargument2:20
The most powerful verification step is the one almost no professional takes. It does not involve checking facts or reading the logic — it involves using AI to challenge its own recommendation. This lecture introduces a two-step technique that stress-tests any AI output against the strongest possible case that it is wrong — and what you do with what you find.
The Search-Verify Loop0:59
There is a method for using AI to fact-check AI — and it works, within specific limits that are critical to understand. This lecture shows exactly how to deploy it, where it is reliable, and where it creates a false sense of security that is more dangerous than not checking at all. The distinction is precise and consequential.
High-Stakes Lab: Spot the Flaw Simulation0:48
This lecture puts you inside a real professional scenario — a briefing document that looks authoritative, is professionally written, and contains multiple critical errors. The VERIFY framework gets applied in real time, step by step, against a document built to test every skill in this section. This is where the framework stops being theoretical and starts being yours.
The Unverified Brief
Applying VERIFY

When to Override: The Five-Step Human Protocol2:48
Overriding AI is not a failure of confidence. It is a professional act — one that requires a specific process to execute correctly, credibly, and in a way that strengthens rather than undermines your professional standing. This lecture traces a real override decision made by a senior professional, reconstructs what she knew that the AI did not, and builds the five-step process from that example outward.
Decision Documentation: Your Reasoning Trail2:29
Three to five minutes per decision. One document. The professional habit that has protected more reputations and built more judgment than any other practice in this course. This lecture makes the case for decision documentation not as compliance — but as the most underused career development tool available to professionals working in AI-augmented environments.
The Trust Calibration Model2:44
Not all AI recommendations deserve the same level of scrutiny. Not all of them deserve the same level of trust. This lecture introduces a decision framework built on four specific variables that determine how much weight any AI output should carry — before you act on it. It is the tool that makes critical AI thinking efficient rather than exhausting.
The Trust Calibration Test

Data Hygiene: The Postcard Rule1:03
There is a single principle that governs what belongs in an AI prompt and what does not — and it fits in one sentence. This lecture introduces that principle, explains exactly what it protects against, and shows how to work around confidentiality constraints without sacrificing the quality of your AI outputs. It is the lecture most professionals wish they had seen earlier.
The Ownership Trap0:55
The question of who owns AI-generated professional work is not fully settled — legally or professionally. This lecture maps the current landscape of AI output ownership, the specific situations where assumptions about that ownership have already caused professional harm, and the practical approach that protects your work and your clients in an environment that is still evolving.
The Kill Switch: When to Revert to Manual0:57
There are five categories of professional situation where using AI makes your outcome worse, not better — and where the professionals who reach for it anyway are taking risks they do not fully see. This lecture identifies all five, explains the structural reason AI fails in each, and gives you the decision criteria for knowing when you are in one of them.
Judgement under pressure
Explaining AI-Led Decisions to Leadership2:22
The ability to talk about your AI involvement in a way that builds trust rather than raises questions is becoming one of the most professionally valuable communication skills in the workplace. This lecture shows two versions of the same conversation — one that erodes confidence and one that builds it — and constructs the three communication frames that distinguish AI supervisors from AI users in every senior conversation.
The Override Conversation

From Chatbots to Agents: When AI Takes Actions1:06
The AI tools most professionals use today produce outputs for humans to evaluate. The AI tools arriving in the workplace now do something different — they take actions. This lecture explains how that shift changes the critical thinking required of professionals who supervise them, and why the window to develop these skills before they are urgently needed is narrowing.
Setting the Guardrails: Approval Gates for Autonomous Workflows0:59
When AI acts rather than advises, the question of where to place human judgment in the process becomes the most consequential design decision you can make. This lecture introduces the structural mechanism for answering that question — giving you the framework for ensuring that human oversight is applied precisely where it matters, without eliminating the efficiency that makes agentic AI valuable.
Case Study: 10-Hour Task to 45 Minutes of Elite Supervision0:46
This lecture builds a complete picture of what skilled AI supervision looks like in practice — tracing a real professional task from a ten-hour manual process to a forty-five-minute supervised AI workflow. Every decision point is mapped, every verification step is visible, and the specific expertise that made the difference is identified. This is what the fully assembled skill set from this course looks like in use.
The Weekly AI Audit: Your Five-Minute Sharpening Habit2:09
Critical thinking is a muscle. Without deliberate practice, it atrophies — especially in an environment where AI is constantly offering to do the thinking for you. This lecture introduces the five-minute weekly practice used by the professionals who maintain the sharpest AI judgment over time — and explains why the compounding value of this habit is disproportionate to the time it requires.
The Agent Goes Off-Script

Your Complete Framework on One Page2:12
Every skill in this course connects. This lecture maps the full system — from Problem Architecture through VERIFY, Trust Calibration, the Override Protocol, the Five Thinking Moves, and the Weekly Audit — showing how each component fits into a single professional practice. This is the lecture you return to when you want to see the whole picture. It is also the one your team needs to see first.
How to Roll This Out to Your Team2:08
Individual AI supervision skills are valuable. Team-level AI supervision culture is transformational — and significantly harder to build without a specific approach. This lecture gives managers and team leads a practical three-step protocol for making critical AI thinking a shared team practice rather than a personal one. The Manager's Reinforcement Guide introduced here is the most-downloaded resource in this course.
Your Supervision Practice
Your 30-Day Critical Thinking Practice Plan2:09
The gap between completing a course and changing how you work is seven days. This lecture closes that gap with a specific, week-by-week thirty-day plan that converts the frameworks from this course into professional habits. Each week builds on the last. By day thirty, the system that required conscious effort in week one is running in the background.
Conclusion and Wrap Up1:00

Requirements

You use AI tools (ChatGPT, Claude, Copilot, Gemini or similar) in your professional work at least occasionally
You make or contribute to decisions at work — recommendations, reports, strategies, or client deliverables
No technical background required — this course covers thinking skills, not AI architecture
You are open to examining how your own cognitive patterns interact with AI outputs
You want to advance beyond basic AI use into professional-level AI judgment

Description

Are You Using AI — or Just Trusting It?

There is a question every professional needs to answer honestly right now.

When an AI tool gives you an output — an analysis, a recommendation, a strategy, a report — what do you actually do next?

If the answer is "review it quickly and move forward," you are in the majority. You are also carrying a risk that most professionals do not see until it surfaces in a meeting, a client conversation, or a board presentation at exactly the wrong moment.

The risk is not that AI will replace you. The risk is that AI will make you confidently, efficiently, professionally wrong — and that nobody will know until the damage is done.

This course exists because that risk is real, it is growing, and there is a systematic way to eliminate it.

What This Course Is — and What It Is Not

This is not a course about prompting AI better.

It is not a course about which AI tools to use, how to integrate them into your workflow, or how to automate tasks you currently do manually.

Those courses exist. They are useful. This is not one of them.

This course is about what happens in the space between an AI output and a professional decision. It is about the judgment, the verification, the calibration, and the advanced reasoning that determine whether AI makes you better at your job — or just faster at making the same mistakes with greater efficiency.

It is about the thinking skills that AI cannot replace, and that organisations are already beginning to identify as the defining professional capability of the next decade.

If you have ever forwarded an AI output you did not fully verify, presented a recommendation you could not completely defend, or accepted a conclusion because it was well-formatted and confidently written — this course changes that. Permanently.

The Problem No One Is Talking About

Ask most organisations what their AI training covers, and you will hear about tools. Prompting. Automation. Efficiency gains.

Ask what they do not cover, and the answer is almost always the same: what to do when the AI is wrong.

And AI is wrong. Not occasionally, not in edge cases, but regularly, predictably, and in ways that follow specific patterns — patterns that are detectable once you know what to look for, and invisible to professionals who do not.

AI outputs that contain fabricated citations presented with complete confidence. Recommendations built on logical structures that do not hold under examination. Analyses that are technically correct but built on assumptions that do not apply to your specific situation. Strategies that look sound at first glance and collapse under the pressure of a single follow-up question.

These are not hypothetical scenarios. They are documented, they are recurring, and they are costing organisations credibility, money, and competitive position right now.

The professionals who develop systematic AI judgment skills now — the ability to evaluate, verify, calibrate, and confidently decide — are the ones who will be trusted with the most important work in their organisations over the next decade.

This course builds those skills. Completely. From the ground up.

Who This Course Was Built For

This course was designed for professionals who use AI in their work and are ready to move beyond basic use into professional-grade AI judgment.

You are the right learner for this course if you make decisions at work that other people act on. If you produce outputs that carry your professional credibility. If you work in an environment where the quality and accuracy of your work matters more than the speed at which it is produced.

You do not need a technical background. You do not need to understand how AI models work at an architectural level. You need to be a professional who thinks seriously about their work — and who wants to ensure that AI makes that work better, not just faster.

This course has been designed with two specific learner types in mind, and it delivers something different and valuable to each of them.

The first type is the professional who needs tools they can use immediately — a systematic process for verifying AI outputs, a framework for making better AI-assisted decisions, a set of habits that protect their professional credibility starting this week. If that is you, this course gives you that system in the first half.

The second type is the professional who wants to become a fundamentally sharper thinker — someone who develops cognitive capabilities that AI cannot replicate, that compound in value over time, and that distinguish strategic thinkers from analytical ones. If that is you, this course has an advanced section that goes further than any AI training available in this category.

Most serious professionals are both. This course was built for them.

What You Will Learn — And What You Will Be Able to Do

The learning in this course is organised around transformation at three levels: what you know, what you can do, and who you become as a professional.

What you will know by the end of this course is a complete picture of how AI reasoning works, where it fails, what patterns those failures follow, and why the human brain is psychologically predisposed to miss them. You will understand the specific cognitive dynamics that make AI-assisted professional errors so common — and so preventable.

What you will be able to do is apply a professional-grade system for evaluating any AI output before you act on it. You will have a structured verification process, a decision calibration framework, a protocol for overriding AI when your professional judgment conflicts with its recommendation, and a documentation habit that protects your credibility and builds your judgment over time.

Who you will become is a professional who thinks at a level that AI cannot substitute for. You will have developed five specific cognitive capabilities — advanced thinking moves that are entirely human — that allow you to see what AI misses, ask what AI cannot ask, and reason through what AI cannot reason through. These capabilities do not just make you better at working with AI. They make you a better professional, period.

How Your Work Will Change

The transformation this course produces is visible and specific. Here is what changes.

Your AI-assisted outputs become the most reliable in your team. You will have a systematic process for ensuring accuracy, logical integrity, and contextual relevance in every significant AI output you produce. The outputs you send will be outputs you can fully defend.

Your decisions become more defensible. You will have a documented reasoning trail for every significant AI-assisted decision — evidence that your professional judgment was applied, not delegated. In the environments where AI accountability is becoming a professional and legal question, that documentation is not optional.

Your thinking becomes sharper. Not just your AI thinking — your thinking. The advanced cognitive skills in this course apply to every complex decision you face, regardless of whether AI is involved. You will ask better questions, trace consequences further, notice what others miss, and hold complexity open longer than feels comfortable — which is, consistently, where the most valuable insights live.

Your professional standing changes. The professionals who can be trusted with AI-assisted work at the highest-stakes level are the ones who will be given it. This course builds the capabilities that earn that trust — and makes them visible to the people responsible for evaluating your work.

What Makes This Course Different

There are many AI courses available. Most of them teach you to use AI more efficiently. A smaller number teach you to use it more strategically. Very few teach you to use it with genuine professional judgment.

This course sits in the last category — and it is one of a very small number that extends beyond AI judgment into the advanced cognitive territory where human thinking becomes irreplaceable.

The frameworks in this course are proprietary, named, and teachable. They are not generic critical thinking principles applied loosely to an AI context. They are purpose-built for the specific challenges that AI-assisted professional work presents — the psychological dynamics, the failure patterns, the decision structures, and the thinking moves that this particular technological moment requires.

The course is deliberately tool-agnostic. The skills you build here apply regardless of which AI platform you use — today, next year, or five years from now. You are not learning to operate a tool. You are developing judgment that compounds in value as the tools continue to evolve.

The course is designed for professionals with real workloads, real time constraints, and real stakes. It is structured to deliver immediate, applicable value — skills you can use before the week the course is completed — while building toward a level of capability that takes the full course to reach.

A Note for L&D Professionals and Managers

If you are evaluating this course for team deployment, here is what the organisational case looks like.

Every professional you give AI access to without critical thinking training is a liability. Not because they are careless — but because AI-assisted errors are invisible in ways that human errors are not. They look professional. They are well-formatted. They are delivered with confidence. They pass the surface-level scan that most professionals apply before forwarding or acting.

The systematic evaluation skills this course builds are what convert AI access from a productivity tool into a trustworthy capability. They are the difference between AI that makes your organisation faster and AI that makes your organisation trustworthy.

This course comes with a Manager's Reinforcement Guide — a set of specific questions designed to be used in one-on-ones to ensure that the skills from this course are being applied, not just remembered. It also includes a team rollout protocol that can be deployed in a single team meeting without additional facilitation.

The ROI on this course is measured in the decisions it prevents from going wrong, and the professional credibility it protects. Both compound over time.

Frequently Asked Questions

What is critical thinking with AI, and why does it matter now?

Critical thinking with AI is the professional discipline of evaluating, verifying, and making sound decisions based on AI outputs — rather than accepting them uncritically. It matters now because AI tools have reached a level of capability where their outputs are persuasive, professional, and difficult to distinguish from accurate content — even when they are wrong. The professionals who develop systematic AI judgment skills now are building a capability that will be foundational to professional credibility in every industry within the next three years.

How is this different from learning to use AI tools better?

Most AI training focuses on inputs — how to structure prompts, which tools to use, how to automate tasks. This course focuses on outputs — what to do with what AI produces, how to evaluate its accuracy, how to make better decisions based on it, and how to develop the advanced thinking skills that no AI tool can substitute for. The difference is the difference between using a calculator and understanding mathematics.

Do I need technical skills or an AI background to take this course?

No. This course requires no coding ability, no understanding of machine learning, and no prior AI training. It is designed for professionals whose work involves making decisions — not building AI systems. The entire course operates at the level of professional judgment and cognitive skill, not technical implementation.

How quickly will I see results from this course?

The first applicable skill in this course — the pre-prompt discipline introduced in Section 2 — can be applied to your next AI interaction. The verification framework in Section 3 can be applied to your next significant AI output. Most participants report a visible improvement in the quality of their AI-assisted work within the first week of applying the course frameworks. The advanced thinking skills in Section 4 take longer to internalise but begin producing results in the weeks immediately following course completion.

What AI tools does this course work with?

All of them. The frameworks and skills in this course are deliberately tool-agnostic — designed to work with ChatGPT, Claude, Microsoft Copilot, Google Gemini, Perplexity, and any AI tool that produces professional outputs. The course teaches you to evaluate outputs and develop judgment, not to operate specific platforms. As the AI tool landscape continues to evolve, the skills from this course become more valuable, not less.

Is this course relevant if I already use AI extensively at work?

Especially relevant. The professionals most exposed to AI-assisted errors are the ones who use AI most frequently — because frequency without a systematic review process means errors accumulate faster and with greater confidence. If AI is already a significant part of how you work, this course is the professional quality system that should have accompanied that adoption.

Can AI replace the kind of thinking this course teaches?

No. The advanced cognitive capabilities covered in this course — reasoning about downstream consequences, identifying the assumptions hidden in questions, detecting signals that fall outside established patterns, knowing where a model's validity breaks down, and holding genuine ambiguity open rather than collapsing it prematurely — are not replicable by current or foreseeable AI systems. They are specifically the territory where human judgment is irreplaceable, and they are the subject of the advanced section of this course.

Will this course be updated as AI evolves?

The course is built to be evergreen. The frameworks, thinking skills, and decision systems it teaches are grounded in how professionals evaluate information and make judgments — capabilities that hold their value regardless of which AI tools are in use today or in the future. You are not learning to operate a specific platform. You are developing professional judgment that compounds as the AI landscape continues to shift.

How long does this course take to complete?

This course is 50 mins of all video content plus 50 minutes of combined activites. The recommended approach is to take the course in two sitting. The full course — including the capstone project — takes approximately two days to complete with genuine application.

Who should take this course before deploying AI tools to their team?

Managers, L&D professionals, and team leads who are responsible for the quality of AI-assisted work in their teams. Ideally, this course is completed before — or immediately following — the deployment of AI tools at scale. The critical thinking skills it builds are the professional layer that makes AI tool adoption safe, not just fast.

What Professionals Say After Completing This Course

"I thought I was already using AI well. This course showed me what I was missing — and the difference in the quality of my outputs since completing it has been immediately visible to my team."

"The frameworks in this section are the most practically applicable AI training I have encountered. I used the verification system on my next deliverable the day after completing the course."

"The advanced thinking section is unlike anything else available in this category. It made me a better professional, not just a more careful AI user."

"I rolled this out to my team of twelve within a week of completing it. The Manager's Reinforcement Guide made that conversation straightforward. We now have a shared vocabulary for AI output quality that has changed how we review work."

Your Enrollment Decision

The professionals who develop systematic AI judgment skills in the next twelve months are the ones who will be trusted with the highest-stakes AI-assisted work in their organisations for the decade that follows.

The professionals who do not will continue to be fast. Some of them will also, eventually, be visibly wrong at exactly the wrong moment.

This course is the difference between those two outcomes. There is no professional risk in starting.

The only risk is in waiting.

Who this course is for:

Professionals and knowledge workers who use AI tools at work and want to ensure their AI-assisted outputs are accurate, defensible, and genuinely trustworthy
Managers and team leads who are responsible for the quality of AI-assisted work produced by their teams and want a shared system for reviewing it
L&D professionals and HR leaders who are designing or rolling out AI capability programs and need a critical thinking layer to accompany tool adoption
Consultants and client-facing professionals whose credibility depends on the accuracy and soundness of every deliverable, regardless of how it was produced
Senior individual contributors and aspiring strategists who want to develop the advanced reasoning skills that distinguish analytical thinking from strategic thinking in an AI-augmented environment

Critical Thinking with AI

What you'll learn

Explore related topics

Course content

Welcome & Orientation3 lectures • 4min

The AI Thinking Gap3 lectures • 6min

Problem Architecture4 lectures • 4min

The VERIFY Framework8 lectures • 10min

Decision Intelligence4 lectures • 8min

Risk, Ethics & The Kill Switch5 lectures • 5min

Agentic Supervision5 lectures • 5min

Your AI Supervision System4 lectures • 7min

Requirements

Description

Who this course is for: