20 "Critical Thinking Prompts" Are Really About 6 Real Techniques
Twenty different-sounding prompt frameworks for critical thinking turn out to be a much smaller number of actual reasoning techniques, most of them decades old, now just packaged for a chat window.
One real cluster is about deliberately arguing against yourself: a devil's advocate stress test, a red team analysis, and a cognitive bias check are all versions of the same discipline -- assigning someone, or something, the explicit job of attacking your own idea before reality does it for free. The practice traces back centuries, formalized in the Catholic Church's advocatus diaboli role during sainthood reviews, and shows up today in military and cybersecurity red teams whose entire job is finding the weakness before an actual adversary does. A second cluster is about structuring a decision rather than just discussing it: a weighted decision matrix, an opportunity-cost evaluation, and formal scenario planning all come from real decision-analysis and strategy methodology -- scenario planning specifically was pioneered by Royal Dutch Shell in the 1970s to prepare for oil-market futures nobody could predict with confidence, and it worked well enough that it became standard practice across corporate strategy.
A third cluster digs into causes rather than symptoms: root cause analysis is most associated with Toyota's "Five Whys" technique, developed on their production line specifically to stop teams from patching a symptom and calling it fixed, and hypothesis testing borrows directly from the scientific method -- state a falsifiable claim, then look for the evidence that would prove it wrong, not just the evidence that confirms it. A fourth cluster is a set of mental models popularized largely through Charlie Munger's writing on multidisciplinary thinking: first-principles thinking (strip a problem back to what is actually true rather than inherited assumptions), second-order thinking (what happens after the first, obvious effect of a decision), and inversion (instead of asking how to succeed, ask specifically how you would guarantee failure, then avoid doing those things).
None of these require an AI assistant to use -- they are just decision-making and reasoning techniques with a long track record, and a chat prompt is only a convenient way to force yourself through the discipline of actually doing one instead of skipping straight to a conclusion. The genuinely useful move common to all of them is the same one: before you settle on an answer, deliberately construct the strongest case against it, using someone else's real reasoning tools rather than just asking an AI to agree with what you already believed going in.
Drawing on Charlie Munger's writing on mental models, the Toyota Production System's "Five Whys" technique, Royal Dutch Shell's origination of corporate scenario planning in the 1970s, and the military/security origins of red-teaming and devil's advocacy.
Get your own AI exposure score
Two minutes, free — see how exposed your specific role is and what to do about it.
Get your AI Vulnerability Score