Clear out the definite no's first, so the ideas worth a closer look can stand out.
A good brainstorm leaves you with more ideas than you can weigh carefully. Scoring all 60 properly would take all afternoon, and by idea 40 nobody is judging well. Negative selection takes a different first step. Instead of hunting for the winner, you quickly remove the ideas that clearly won't work, and keep everything else.
Each idea goes into one of two piles: No or Maybe. There is no Yes pile yet, which keeps the sort fast and stops early favorites from crowding out the rest. One rule protects the good ideas: when in doubt, it's a Maybe. Repeat the sort, with more care on each pass, until the Maybe pile is short enough for a closer look.
Use it when you have a long list after a brainstorm, when careful scoring would take too long, or when you need a manageable shortlist before you use a finer tool.
A family's 38 birthday ideas became a shortlist of six, with two bold ideas rescued along the way.
Choose a part of the diagram or a tab to explore each step.
What is this list for, and what would make an idea a clear no?
Is this idea clearly unworkable, or worth another look?
Did any surprising idea get thrown out just because it felt unfamiliar?
Looking more closely, which Maybes are really No's?
Which finer tool will you use to choose among the survivors?
Negative selection is how most of us sort a pile of mail. Doing it deliberately has real advantages.
Spotting a clear flaw takes seconds. Deciding which of two good ideas is better can take an hour. By making only the easy calls first, you shrink the list fast and save your careful thinking for the ideas that deserve it.
There's no Yes pile, so no idea gets crowned too soon. A long list often hides a strong idea near the bottom. Keeping everything that isn't a clear no gives late arrivals and quiet ideas a fair chance.
Judging dozens of ideas carefully wears people out, and tired judges fall back on safe, familiar picks. Quick passes keep energy high for the final choice, when it matters most.
Filters tend to strip out anything unusual, leaving a shortlist that is sensible but forgettable. The “when in doubt, keep it” rule, plus a deliberate rescue check, makes sure some daring ideas survive to be judged properly.
Reread the problem and note the few true deal-breakers, such as budget limits, legal rules or values. Agree that “unusual” and “needs work” are not reasons for a No.
Go through the list quickly and put each idea in No or Maybe. Make each call in seconds. If you hesitate, it's a Maybe. Don't discuss, improve or combine ideas yet.
Scan the No pile once. Bring back anything surprising, fixable or that someone feels strongly about. Then check that the Maybe pile still holds a few bold ideas.
Sort the Maybe pile again, reading each idea more carefully and merging near-duplicates. Repeat until about three to eight ideas remain.
Pass the shortlist to a finer tool such as a NUF Test, a Hundred Dollar Test or a PINC Filter. Keep the No pile; rejected ideas sometimes become useful later.
Your ideas on cards or sticky notes, two areas labeled No and Maybe (or two spreadsheet columns) and 15–30 minutes. With AI, use the prompt below.
Goal: a celebration everyone from age 4 to 80 can enjoy. Deal-breakers: over $1,500, more than two hours' travel for Grandma, anything she'd find fussy. Sorters: three grandchildren.
Under $1,500, short travel for Grandma, nothing too formal. “Silly” and “never been done” were not allowed as reasons for a No.
Out went the cruise, the surprise flight to Italy and the black-tie gala. Twenty-three No's in ten minutes left 15 Maybes.
Two ideas came back: an evening of old family photos and stories, and a recipe book of her dishes written by the whole family.
Two more passes, reading each idea aloud. The lake picnic and the park barbecue merged into one. Seventeen became six.
Lake picnic, recipe-book reveal, photo evening, garden tea party, boat ride and a family talent show, ready for a Hundred Dollar Test.
Both rescued ideas made the final six, and the recipe book became the centerpiece of the day.
AI can sort a long list in seconds, which makes it a handy first-pass helper. The danger is that it sorts toward the safe and the familiar. Give it your deal-breakers, ask it to explain every No, and make it argue for the strange ideas. You make the final calls, especially the close ones.
Get a quick sort, but insist on a reason for every No so you can overrule it.
Sort these 38 ideas into No and Maybe using only these deal-breakers. Give a one-line reason for every No.
Make the AI argue for the ideas it rejected before you finally let them go.
Make the best case for the three most unusual ideas in the No pile. Could a small change save any of them?
Combine near-duplicates so the next pass has fewer, stronger ideas.
Which ideas on my Maybe list are really the same idea? Merge them and give each merged idea a clear name.
Check whether your shortlist has lost all its daring ideas.
Rate my shortlist from safe to bold. If it's all safe, which rejected idea would you bring back?
Paste your list, pick a mode, then copy the prompt into Claude, ChatGPT, Gemini or Copilot. In coach mode the AI helps you agree your deal-breakers, sorts with you pass by pass, and never discards anything without asking.
The “Open in” links put the prompt in the web address. If your challenge includes anything sensitive, copy and paste instead.
After the session, click one to copy it, then paste it into the same chat.