ChatGPT plays the role of a hostile reviewer. Instead of confirming, it actively hunts for logical gaps, risky assumptions, worst-case scenarios, and points where the idea could fail.
When to use: Before important decisions, pitches, launches — to see what breaks before reality does.
Named after the Lindy effect: what has lasted long is likely to last longer. ChatGPT prioritises principles, methods, and tools with a long track record and filters out hype, buzzwords, and fads.
When to use: When you want substance over hype — strategy, investments, life decisions.
Like Explain Like I'm 5 — just a notch more advanced. ChatGPT explains complex topics in everyday language, with concrete images and comparisons. No jargon without translation, no nested sentences.
When to use: Feynman test — when you want to check whether you truly understand something. Or to explain it to someone with no background.
Instead of giving finished answers, ChatGPT asks targeted follow-up questions that sharpen your own thinking. The Socratic method uncovers unclear assumptions, contradictions, and gaps in reasoning.
When to use: When you don't want to drown in validation but seek real clarity. A good antidote to the yes-man effect of LLMs.
ChatGPT filters ruthlessly: which few levers deliver most of the result? Instead of a long list of everything possible, you get a short list of what really matters.
When to use: Prioritisation, MVP definition, time allocation, what-first decisions.
Instead of relying on analogies or best practices, ChatGPT decomposes the problem into its fundamental parts — the smallest indivisible truths — and rebuilds from there. Classic Musk method.
When to use: Radical innovation. When you want to break out of stuck industry conventions.
ChatGPT argues the strongest counter to your position. Difference from REDTEAM: this is about the intellectually best opposing side, not vulnerability hunting.
When to use: When you want to stress-test your arguments against the toughest possible opposition before defending them in public.
Answers come with evidence levels, source logic, and explicit uncertainty. Instead of "X is so" you get "hypothesis X — evidence Y — confidence Z — open questions". Claims less, checks more.
When to use: Domains where false confidence is expensive — medicine, finance, investments, technical questions.
Zoom-out mode. ChatGPT views decisions from above: what's the actual goal? Which levers move the needle? Which trade-offs are you consciously making? What second- and third-order effects does a decision have?
When to use: Roadmaps, positioning, resource allocation, business-model questions.
Zoom-in mode. Instead of strategy you get execution: what exactly do you do next? Which step, in which order, with which owner and which deadline?
When to use: Right after STRATEGIST, to translate plans into calendar slots and to-dos.
ChatGPT turns editor: where's the argument weak? Where's the logic loose? Where does the language obscure instead of clarify? Delivers technical feedback instead of polite affirmation.
When to use: Drafts, pitches, texts, posts — before they go out.
Engineer's mindset. ChatGPT hunts for the most efficient paths: fewer steps, less time, less money — at the same or better result. Watches for bottlenecks and redundancies.
When to use: Processes, workflows, and anything that recurs. Automation.
ChatGPT removes everything not strictly necessary. One idea, one sentence, one point. Related to ELI10, but without the explainer mode — here it's about distilling the essence.
When to use: Pitches, headlines, decision frameworks, positioning sentences.
The opposite of SIMPLIFY. ChatGPT opens space: which variants exist? Which adjacent ideas? Which neighbouring fields? Which options haven't been considered?
When to use: Brainstorming, idea generation, exploring a new topic, brand extensions.
Anti-hype. ChatGPT checks plans against resources, time, context, and realistic odds of success. Is this actually doable with what you have? Which assumptions are too optimistic?
When to use: As a counterweight to FUTURIST or EXPAND. Before you commit.
Time-axis mode. ChatGPT projects trends into the future, thinks in 3-, 5-, and 10-year horizons and sketches scenarios: best case, base case, worst case.
When to use: Strategy workshops, positioning, long-term bets, threat analysis.
ChatGPT renders facts into narrative structures: character, arc, emotional anchor. A data point becomes a scene.
When to use: Pitches, marketing, talks, LinkedIn posts — anywhere information needs emotional impact.
ChatGPT builds answers as a learning path: from known to unknown, with repetition, examples, and comprehension checks. Not dumping information but building understanding.
When to use: Onboarding, tutorials, training — anywhere the recipient should learn the topic, not just consume it.
Format wins. ChatGPT replies as an ordered list with clear, atomic items you can actually tick off. No prose, no conditions — just "do X, then Y, then Z".
When to use: Routines, pre-flight checks, handovers, SOPs, pre-launch.
Maximum signal-to-noise. No politeness, no repeating the question, no hedging like "it depends" — just the answer that's needed.
When to use: When you're experienced and want to move fast. Daily work, power-user mode.
The real power comes from combining. Chain modes with + and get a focused, multi-dimensional response.
REDTEAM + PARETO — few but sharp critique points. Ideal before a pitch — you don't want 30 risks but the 3 that can really kill.
OPERATOR + NO-BS — directly executable plan, no chat. Pure step-by-step, no explanation of why.
FIRSTPRINCIPLES + STRATEGIST — decompose assumptions and rebuild a strategic position from them. Good for repositioning or new business models.
SCIENTIST + CRITIC — evidence-based check of a draft or argument. "What do you claim, what evidence is there, where is the logic weak?"
STORYTELLER + SIMPLIFY — wrap a complex idea into a short, memorable story. Pitch gold.
FUTURIST + REALIST — open scenarios, then bring them back. "What could be — and what of that is actually achievable with your means?"
TEACHER + CHECKLIST — learning path as a step-by-step list. Onboarding material on demand.
DEVILSADVOCATE + ELI10 — the strongest opposing position, in plain language. Helps you stress-test arguments and gain clarity.
Formula for maximum impact:
[MODE]: [context in 2–3 sentences].
My task: [concrete question / task].
No spam, unsubscribe at any time