Significance and innovation, without inflation
Lesson 2 of 6 · Grants, talks and posters
1 · Learn the move · Red-team audit
Inflation is the cheapest kill a reviewer has: one 'first-ever' they can falsify from memory, one 'transformative' with nothing under it, and the section's credibility - and yours - goes with it. The audit runs before internal review does: every superlative and priority claim flagged ('first', 'novel', 'critical barrier', 'paradigm', 'only approach'), and each flag resolved one of three ways - supported by evidence you actually hold, with the citation and its verification state; weakened to what the evidence supports; or cut. Specific and hedged outscores grand and unsupported, because study sections are staffed by exactly the people who know your field's actual history. The audit does not decide whether your work is significant - it makes every sentence about significance carry its own weight.
Red-team my significance and innovation sections: [paste]. Flag every priority or superlative claim: first, novel, only, critical barrier, transformative, paradigm. For each flag, output: the claim / what evidence would support it / what I actually hold (from my materials: [paste]) / verdict - SUPPORT (with citation + verification state), WEAKEN (draft the weaker sentence), or CUT. Do not add new claims. Grand and unsupported loses to specific and hedged.
2 · Your turn. You write the prompt
Your significance section came back from a co-investigator with 'punchier now!' and six new superlatives. Two 'first demonstration' claims are in there, and you can only defend one of them - and only for the specific system, not in general. Study section includes the two labs most likely to know the counterexamples.
Remember: the AI sees only your prompt, not this page. If the situation isn't in your prompt, it doesn't exist.
Optional. These shape the output when you run your prompt below, not your score.