Nice! If you end up with a predictive component, I—and perhaps my forecasting group—would be up for giving probability estimates of their chosen metric of success, and maybe estimates of their value. I imagine that these forecasts/estimates would at first be shitty, since they would be in a new domain, and it’s hard to get intuitions, or baserates & distributions of success. But I’d be up for trying!
One suggestion we got at the EA Econ retreat was to have a prediction market somehow tracking the expected value paired to each certificate as a requirement so that the history of the expected value is known, reducing the risk of retro funding ex ante risky projects.
Nice! If you end up with a predictive component, I—and perhaps my forecasting group—would be up for giving probability estimates of their chosen metric of success, and maybe estimates of their value. I imagine that these forecasts/estimates would at first be shitty, since they would be in a new domain, and it’s hard to get intuitions, or baserates & distributions of success. But I’d be up for trying!
See also: I will bet on your success on Manifold Markets, which might be of interest to readers more generally.
One suggestion we got at the EA Econ retreat was to have a prediction market somehow tracking the expected value paired to each certificate as a requirement so that the history of the expected value is known, reducing the risk of retro funding ex ante risky projects.
I imagine that this would at first take too much time and produce somewhat low-quality predictions.