Possibly, prizes can be directed to specific proposals. I can imagine a platform (which could be the comment section of a post) where people write what they intend to write about, or what they’d want someone else to write about, and people can comment and upvote on what seems to them important (and why). Out of these, several such proposals can be chosen and offered a bounty on them—which would be the maximal bounty if a thorough and well-written post would be written on them. Then, quality and how much the original question was solved would be assessed and the prize given would be the bounty times the precentage of the question solved times some quality factor.
I really like this idea. CEA could offer people they trust to use the Forum Prize money to announce prizes for exploring a topic or specific question and writing a post about it.
Why this might be a good idea
1. incentivize useful content (and others could chime in and add to the prize pool, signalling higher usefulness)
2. direct attention for people looking for useful work they might do as a sideproject
Some possible downsides
1. Draws in people that are not primarily altruistically motivated
could be an upside, too, if they do great work and get drawn in, as I expect most smart people will be happy to be able to contribute to improve the world
2. Disagreements about fulfilling the goal criteria might cause frustration
3. Increase the noise in the forum, i.e. more posts of lower quality
Skimming a bit through the forum, this idea has been tried a few times! But less often then I’d have expected if this were a good idea. Some examples:
I just found out that there is a tag on LW for active and closed bounties, listing 4 and 24 prizes in total. My impression is that experimenting with bounties on the EA forum would be a good idea and I wonder what others here think about it.
This can cause perverse incentives. People can hold back on a written post until a bounty is made for that post. People would disproportionally compete on working on the selected issues. People might be discouraged from starting a post they know they won’t get a prize for
Possibly, prizes can be directed to specific proposals. I can imagine a platform (which could be the comment section of a post) where people write what they intend to write about, or what they’d want someone else to write about, and people can comment and upvote on what seems to them important (and why). Out of these, several such proposals can be chosen and offered a bounty on them—which would be the maximal bounty if a thorough and well-written post would be written on them. Then, quality and how much the original question was solved would be assessed and the prize given would be the bounty times the precentage of the question solved times some quality factor.
I really like this idea. CEA could offer people they trust to use the Forum Prize money to announce prizes for exploring a topic or specific question and writing a post about it.
Why this might be a good idea
1. incentivize useful content (and others could chime in and add to the prize pool, signalling higher usefulness)
2. direct attention for people looking for useful work they might do as a sideproject
Some possible downsides
1. Draws in people that are not primarily altruistically motivated
could be an upside, too, if they do great work and get drawn in, as I expect most smart people will be happy to be able to contribute to improve the world
2. Disagreements about fulfilling the goal criteria might cause frustration
3. Increase the noise in the forum, i.e. more posts of lower quality
Skimming a bit through the forum, this idea has been tried a few times! But less often then I’d have expected if this were a good idea. Some examples:
Call for: Essay on “How to evaluate small-scale giving opportunities”, prize: 300$, result: no entries in time (two commenters signalling interest)
Call for: Arguments/comments for or against donating to the EA Hotel, prize: 100$, result: 35 comments, looks good
Call for: Arguments/comments if psychedelics are a potential EA cause area, prize: 400$, 200$, 100$, result: 114 comments, looks good, too
Call for: “Spaced Repetition literature review” on LW, prize: 385$, result: 2 reviews, Gwern won (and the prize seemingly motivating him to write it)
Call for: Essay Contest: Cryonics and Effective Altruism, prize: 1BTC (then 350$), result: 5 essays, all rated positively
I just found out that there is a tag on LW for active and closed bounties, listing 4 and 24 prizes in total. My impression is that experimenting with bounties on the EA forum would be a good idea and I wonder what others here think about it.
This can cause perverse incentives. People can hold back on a written post until a bounty is made for that post. People would disproportionally compete on working on the selected issues. People might be discouraged from starting a post they know they won’t get a prize for