Question · Evidence · Decide
Learn Critical Thinking
Critical thinking is the disciplined practice of clarifying claims, examining evidence and assumptions, testing alternatives, and revising judgment in proportion to the consequences.
Builds your Critical Thinking and AI Fluency capability.
Start here · one first move
Select a consequential claim from work, media or an AI output.
Field guide 01
Understand
A working definition
In practice
Asks what is being claimed, checks how we know, searches for missing perspectives and explains why the conclusion is warranted.
Why this matters now
In an AI-rich environment, fluent explanations can hide weak premises, missing data or invented sources. Critical thinking keeps speed subordinate to evidence and accountability.
It is
- Purposeful and evidence-sensitive.
- Both a skill and a willingness to question.
- Proportional to consequence.
- Open to revision.
It is not
- Automatic skepticism.
- Winning an argument.
- Finding faults without alternatives.
- Outsourcing verification to a confident tool.
The capability loop
A disciplined judgment loop.
Strong judgment is built through a sequence—not a personality type.
- 1
Frame
What exactly is the claim, decision or problem?
- 2
Decompose
Which assumptions, terms and causal links need separating?
- 3
Verify
What evidence supports the claim, and how credible is it?
- 4
Challenge
Which alternatives, counterexamples or harms are missing?
- 5
Decide
What is justified now, and what would change the judgment?
Go deeper
The full field guide
7 connected capabilities+
Reasoning becomes reliable when its parts can be examined.
- Information analysis
- Breaks complex material into meaningful parts.
- Evidence evaluation
- Assesses credibility, relevance and sufficiency.
- Logical reasoning
- Tests whether conclusions follow from premises.
- Assumption identification
- Makes hidden beliefs and constraints visible.
- Perspective analysis
- Compares interpretations and affected viewpoints.
- Inference drawing
- Forms proportionate conclusions from evidence.
- Metacognition
- Monitors bias, confidence and the quality of reasoning.
The components strengthen one another. Development depends on how they are combined in context—not on maximizing one behavior in isolation.
Behaviour map+
01
Before
- Clarify the question Separates the real decision from the first framing.
- Set evidence needs Names what must be known before acting.
- Map consequence Identifies who may be affected and how.
- Notice incentives Checks whose interests shape the information.
02
During
- Trace claims Connects assertions to sources and methods.
- Test assumptions Looks for hidden premises and boundary conditions.
- Seek disconfirmation Searches for evidence that could change the view.
- State uncertainty Separates known, inferred and unknown.
03
After
- Explain the judgment Shows how evidence and trade-offs led to the choice.
- Record conditions Names what would trigger review or reversal.
- Check outcomes Compares expected and actual effects.
- Update the method Improves future questions, evidence and gates.
Case story · composite+
When the AI report recommends closing a market.
Critical thinking did not slow the decision.
The challenge.
A leadership team receives a polished AI-generated analysis recommending a fast market exit.
The narrative is coherent, the charts look convincing and the deadline is close. 1 · Frame They restate the decision, stakeholders and cost of a false conclusion. 2 · Decompose The team separates market demand, unit economics, capability gaps and timing. 3 · Verify Two key citations cannot be traced and one comparison uses mismatched periods. 4 · Challenge A local team supplies a counterexample and a lower-risk partnership option. 5 · Decide and learn Leadership pauses the exit, runs a bounded test and documents verification gates.
Capability in practice Critical thinking did not slow the decision.
It prevented speed from disguising weak evidence.
Leadership and learning culture+
- Model Show how your own conclusions change when evidence changes.
- Slow Create deliberate pauses before high-consequence commitments.
- Diversify Invite dissent, domain expertise and affected perspectives.
- Verify Set source, calculation and human-review gates for AI-supported work.
- Learn Review decision quality—not only whether the outcome happened to be good.
Why we protect this skill — HI · FRI · IMPACT+
Human Core
How central should human agency, judgment and responsibility remain?
uman leads
▰▰▰▰▱
Future Readiness
How strongly does this help people learn and act as roles and technology change?
Foundational
▰▰▰▰▰
Safeguard Value
How strongly does this protect capability from over-reliance, deskilling and loss of agency?
Critical safeguard
▰▰▰▰▰
They describe why the capability matters—not how advanced a learner is. AI can generate options and counterarguments. The person must set the question, verify material claims and own the consequential judgment.
Practice
THINK before you commit.
Use this short structure in real work, reflection or coaching. Keep the action specific enough to observe.
- T
Target
What claim or decision are we actually testing?
- H
Hunt
Which assumptions, incentives and missing voices matter?
- I
Inspect
How credible, relevant and sufficient is the evidence?
- N
Name
What alternatives, uncertainties and consequences remain?
- K
Keep / revise
What judgment is justified, and what would change it?
One-minute start: “The claim is … The strongest evidence is … The missing check is …” The protocol supports practice. It does not remove the need for context, expertise, ethical judgment or appropriate human support.
Seven-day experiment
Run one real claim audit.
0 of 7 days marked · saved on this device
Prove it
Five stages of critical judgment.
The progression describes increasingly independent, contextual and socially enabling behavior—not intelligence or personal worth.
L1
Notice
Distinguishes claim from opinion and asks basic source questions with support.
L2
Question
Names assumptions, seeks relevant evidence and explains a simple reason.
L3
Evaluate
Compares sources, tests logic and calibrates conclusions to evidence.
L4
Integrate
Weighs uncertainty, perspectives and consequences across complex contexts.
L5
Enable
Makes reasoning visible, improves decision systems and helps others challenge safely.
Use the progression well Ask for the next observable behavior—not a label for the person.
Ask Markku-AI about Critical Thinking
Pick a question. Markku-AI answers from the Skills Pathfinder material.
Read the full minibook
Critical Thinking
The printed A5 field guide — same content, typeset for paper.