A thought experiment replaces gut-feel decision-making with a disciplined “what if” that runs before you spend money or time. The most useful business thinking tools — pre-mortems, inversion, second-order effects — are simply structured ways to attack a plan from angles your optimism would otherwise skip. This generator pairs your specific challenge with the right mental model and assembles a prompt that makes an LLM run the experiment rigorously.
How it works
You write the challenge — a decision, a bet, or a stuck point — and choose a mental model. A pre-mortem imagines the project has already failed and works backwards to the causes. Inversion asks how you would guarantee the worst outcome, then avoids those moves. Second-order effects trace the consequences of consequences. The tool also offers “apply several,” which instructs the model to run multiple lenses and reconcile what they reveal. The generated prompt tells the model to stay concrete, quantify where possible, and finish with prioritised next actions. Everything is assembled in your browser.
The mental models explained
Pre-mortem
You imagine it is six months from now and the project, plan, or decision has failed. Working backwards, you ask: what went wrong, and why? The pre-mortem is most useful immediately before a commitment, when you are optimistic and least likely to look for failure modes on your own. The generated prompt forces the model to list at least five specific failure causes and rank them by likelihood, so the output is actionable rather than generic.
Inversion
Instead of asking how to succeed, inversion asks: how would I guarantee the worst possible outcome? Then you avoid those moves. For a product launch, for example: “How would I ensure no one buys?” often surfaces pricing, distribution, or messaging problems faster than conventional planning. The prompt instructs the model to be concrete about the inverse, not abstract.
Second-order effects
Most decisions have obvious first-order consequences that everyone plans for. Second-order effects are what happens next — the downstream ripple through people, teams, markets, and systems. For example: cutting a customer success team saves cost (first order) but raises churn (second order), which lowers lifetime value and makes CAC harder to justify (third order). The prompt traces this chain two or three levels deep.
Steelman
The steelman (opposite of a straw man) builds the strongest possible case for a position you doubt or oppose. It is useful when you are dismissing an idea too quickly and want to pressure-test whether the dismissal is right. The prompt asks the model to defend the position rigorously, not find a comfortable middle ground.
Tips and examples
Make your challenge statement specific and decision-shaped — “Should we drop our free tier next quarter to improve margins?” yields a far richer experiment than “thinking about pricing.” For high-stakes calls, run a pre-mortem first to expose failure modes, then run a second-order-effects pass to check the fixes do not create new problems. Treat the output as a hypothesis generator, not an oracle: the value is in the non-obvious risks and consequences it surfaces, which you then verify against your own knowledge. Save the prompts you find most useful and reuse them as a standing decision ritual before any irreversible commitment.