Write a research prompt (for an AI or a person) that tests the requester's question without slipping in the writer's expected answer: keep the requester's goal and scope in their own words, find every term that presumes a result (a tradeoff, a limit, a fixed number of dimensions, a physical constraint, a frontier), delete it or turn it into something to be tested, keep only the constraints the requester gave, and log every change.
Nothing: your AI starts it with your words.
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this workflow by what it does and runs it.
Say to your AI: “Use Trillion to write a research prompt (for an AI or a person) that tests the requester's question without slipping in the writer's expected answer: keep the requester's goal and scope in their own words, find every term that presumes a result (a tradeoff, a limit, a fixed number of dimensions, a physical constraint, a frontier), delete it or turn it into something to be tested, keep only the constraints the requester gave, and log every change”
What your AI calls: start_workflow {"workflow_id":"workflow:write-a-research-prompt-without-the-answer-in-it"}