They should use a few of the sats and do an api call to ppq.ai or routstr and ask if the guess is equivalent enough to the correct answer, and if so, pay out, if no, mark it as wrong.
I love this. I'd like to see this structure applied to more things.
I also wonder if it can be applied to things where the system doesn't know the answer; an answer that's hard to generate but easy to verify. That probably only makes sense if participants can't fully verify it themselves though.
A blur slider on the past photos to show what they looked like as they came into focus would be a fun feature.
These results look scammy because one or more 'wrong guesses' were correct answers.
lol
They should use a few of the sats and do an api call to ppq.ai or routstr and ask if the guess is equivalent enough to the correct answer, and if so, pay out, if no, mark it as wrong.
I love this. I'd like to see this structure applied to more things.
I also wonder if it can be applied to things where the system doesn't know the answer; an answer that's hard to generate but easy to verify. That probably only makes sense if participants can't fully verify it themselves though.
It has the feel of a kind of dutch auction. You could almost tell a story this way: readers make predictions about what happens next.
Hey, I guessed 'rodent' and 'sitting mice' and the answer was just 'mice'. I think that should count.
looking at past rounds, the remote control image had a correct answer of remote controller, but claims it expired without a winner...?
Log Cabin
Cool!