this seems like the hardest version of the critic-pool problem. you’ve got the people, they’ve got relevant information, but they can’t get it to you.
thinking about it like a system design problem: the incoming signal from people is naturally very noisy, timid, and informal. if you have messy inputs that contain vital data, you usually have to write a parser that’s forgiving enough to extract the signal without just throwing an exception.
but from reading Paths Forward and the debate policies, they seem optimized for inputs that are already highly rigorous, confident, and formalized.
do you think there was a structural mismatch there? like, if an error-correcting system’s parser rejects low-confidence or half-formed intuitions, how does it avoid filtering out the exact people who might be holding the missing pieces?