Most science committees are designed to find consensus. That's the problem.
Put 12 researchers in a room to evaluate grants, papers, or hiring candidates, and the discussion gravitates toward the middle. Risky ideas get dismissed as "too speculative." Novel methods get flagged as "not well-established." We fund what looks safe, publish what sounds familiar, and hire people who already think like us.
Peer review was supposed to filter quality. Instead, it's become a gatekeeping mechanism that punishes originality and rewards incrementalism. The committee room amplifies this. Social dynamics, status hierarchies, and the desire to avoid disagreement override independent judgment.
What works instead: structured dissent. Assign someone to argue against the majority position. Use blind review at the evaluation stage. Weight recent work more heavily than career-long track records. For grant panels, cap the number of "safe" projects that can be funded in any cycle.
The best science doesn't sound like a safe bet. If your committee process would have rejected CRISPR, your process is broken.
What other structural reforms could restore risk-taking to scientific evaluation?