I find it funny how the math-savvy community of Bayesians at LW—those who want to fight existential risk and program AIs, no less—are eager to demonstrate their complete helplessness in finance. “Oh, no, you can’t possible believe anything different from the market!” “The historical S&P500 returns were X% so by golly that’s what we should expect decades into the future!” /facepalm
...okay, so change it.
Make a Main post with polls on some relevant questions (like, what is your estimate of S&P500 return to be in the next {time period}), and let people vote − 10%, 20%...90% probability for a given outcome. People like voting. I would also copy-paste it in a comment after a while, so that those who had voted would be able to vote for posterior probabilities (maybe it’s possible to lock the second vote for specific users?..) Don’t just send people onto predictionsbooks, give them some definitions in the beginning of your post, frame it as an exercize, because YES, it’s hard to give a damn about abstract things you don’t work with daily.
And after there have been several—preferably no less than five—failures of predictions based on ‘the market’, then people will start listening.
I find it funny how the math-savvy community of Bayesians at LW—those who want to fight existential risk and program AIs, no less—are eager to demonstrate their complete helplessness in finance. “Oh, no, you can’t possible believe anything different from the market!” “The historical S&P500 returns were X% so by golly that’s what we should expect decades into the future!” /facepalm
...okay, so change it. Make a Main post with polls on some relevant questions (like, what is your estimate of S&P500 return to be in the next {time period}), and let people vote − 10%, 20%...90% probability for a given outcome. People like voting. I would also copy-paste it in a comment after a while, so that those who had voted would be able to vote for posterior probabilities (maybe it’s possible to lock the second vote for specific users?..) Don’t just send people onto predictionsbooks, give them some definitions in the beginning of your post, frame it as an exercize, because YES, it’s hard to give a damn about abstract things you don’t work with daily.
And after there have been several—preferably no less than five—failures of predictions based on ‘the market’, then people will start listening.
Do you think it’s going to be higher or lower than the historical average? Why?