Would a transparent idea directory enable refinement of good ideas into great ones, help great ideas find a team, all the while reducing the overall burden of transaction costs associated with considering new ideas?
A transparent idea of proposals should have some effect in this direction. I’ve asked for a transparent directory of projects for months; it’s something I’d like to see funders like EA Grants and thought-leaders like 80,000 work on. However, we need to be cautious because pure ideas are not very scarce. They may be 20% of the bottleneck but 80% is getting talented people. So new project proposals should be presented in such a way that founders will see these ideas and notice if they are a good fit for them.
I- Ready for implementation. These are extremely well considered ideas that support EA principles and have/will contribute good evidence for effectiveness.
II- Worth refining. These are promising ideas that can be upgraded to type I with more background research, adjustments in strategy, etc.
III- Back to the drawing board. These are well intentioned but miss the mark in an important way, perhaps an over-reliance on intuition or misinformation.
I guess that (II-III) are more like forum posts and should usually be filtered out without need for formal review. I think even most proposals in category (I) are too weak to be likely to suceed. I would use a more stringent checklist e.g. (a) funding may be available (b) part of a founding team is available (c) there is some traction demonstrated.
Too many ideas and not enough doers increases the likelihood that doers will settle on weak ideas… if the number of doers is saturated, they only gum up the works.
There are forces in both directions. If more high-quality ideas are shared, then doers may be less likely to settle on weak ideas.
Finally, the main goal of a transparent idea directory is to reduce the unavoidable transaction costs of new ideas.
Then the focus of such a project should not just be to archive ideas, it should be to have more ideas turned into action.
General thought: I think the quality of ideas is far more important than quantity here. I would much rather see two ultra-high-quality proposals online in a system like this than ten mid-range quality ones. It would be good if people could be encouraged to solicit line-by-line feedback by putting their proposals in google docs, and also if there was a requirement for authors to allow anonymous private feedback. Proposals that are substantially downvoted should perhaps disappear for redrafting. Perhaps team-members should be able to submit themselves as candidates for future projects, awaiting a suitably matched project, IDK. It seems like an important space!
Thanks for the thoughtful comments, agree almost completely, particularly your closing points.
My main quibble is the comparison of talent vs ideas as a bottleneck, where you say talent is 80% of the problem compared to ideas at 20%. I certainly agree that lots of weak ideas pose problems, but the trouble with with this comparison is that the first step to recruiting more talent will be an idea. So, in a sense, the talent gap IS an idea gap. In fact, aside from blind luck, every improvement on what we have will first be an idea. Perhaps we shouldn’t think of ideas in opposition to anything, but instead work to maximize them (and keep the bad ones out of the way). Every gap has an idea component, essentially waiting for a better idea for how to close it.
Additionally, having high-yield, impactful ideas on hand that will make a difference could attract talent that might otherwise see EA as a bunch of airy headed idealists. Finally, if talent rather than ideas is the true bottleneck, then it’s all the more important to make sure talent gets connected with the best ideas.
Minor point- Regarding weak ideas, I think there is some value for people to see (a) what makes bad ideas bad and (b) whether or not a particular idea has already been floated, thereby cutting down on redundancy.
If we’ve got more ideas than talented people, it seems like being able to prioritize ideas well is very important. Having people work on the best 10% of ideas is much better than having people work on 10% of the ideas that are chosen at random. I think a transparent idea directory could be very valuable for prioritization, because in order to prioritize well, you would like to be able to brainstorm as many possible pros and cons for any given idea as possible.
To put it another way, getting more people involved in prioritization means we can take advantage of a broader array of perspectives. See Givewell on cluster thinking. I think having people brainstorm ways in which a proposal might end up being unexpectedly harmful could be especially valuable.
General thought: I think the quality of ideas is far more important than quantity here.
My impression is that idea production is like pottery in the sense that the best way to get quality is to aim for quantity.
A transparent idea of proposals should have some effect in this direction. I’ve asked for a transparent directory of projects for months; it’s something I’d like to see funders like EA Grants and thought-leaders like 80,000 work on. However, we need to be cautious because pure ideas are not very scarce. They may be 20% of the bottleneck but 80% is getting talented people. So new project proposals should be presented in such a way that founders will see these ideas and notice if they are a good fit for them.
I guess that (II-III) are more like forum posts and should usually be filtered out without need for formal review. I think even most proposals in category (I) are too weak to be likely to suceed. I would use a more stringent checklist e.g. (a) funding may be available (b) part of a founding team is available (c) there is some traction demonstrated.
There are forces in both directions. If more high-quality ideas are shared, then doers may be less likely to settle on weak ideas.
Then the focus of such a project should not just be to archive ideas, it should be to have more ideas turned into action.
General thought: I think the quality of ideas is far more important than quantity here. I would much rather see two ultra-high-quality proposals online in a system like this than ten mid-range quality ones. It would be good if people could be encouraged to solicit line-by-line feedback by putting their proposals in google docs, and also if there was a requirement for authors to allow anonymous private feedback. Proposals that are substantially downvoted should perhaps disappear for redrafting. Perhaps team-members should be able to submit themselves as candidates for future projects, awaiting a suitably matched project, IDK. It seems like an important space!
Thanks for the thoughtful comments, agree almost completely, particularly your closing points.
My main quibble is the comparison of talent vs ideas as a bottleneck, where you say talent is 80% of the problem compared to ideas at 20%. I certainly agree that lots of weak ideas pose problems, but the trouble with with this comparison is that the first step to recruiting more talent will be an idea. So, in a sense, the talent gap IS an idea gap. In fact, aside from blind luck, every improvement on what we have will first be an idea. Perhaps we shouldn’t think of ideas in opposition to anything, but instead work to maximize them (and keep the bad ones out of the way). Every gap has an idea component, essentially waiting for a better idea for how to close it.
Additionally, having high-yield, impactful ideas on hand that will make a difference could attract talent that might otherwise see EA as a bunch of airy headed idealists. Finally, if talent rather than ideas is the true bottleneck, then it’s all the more important to make sure talent gets connected with the best ideas.
Minor point- Regarding weak ideas, I think there is some value for people to see (a) what makes bad ideas bad and (b) whether or not a particular idea has already been floated, thereby cutting down on redundancy.
If we’ve got more ideas than talented people, it seems like being able to prioritize ideas well is very important. Having people work on the best 10% of ideas is much better than having people work on 10% of the ideas that are chosen at random. I think a transparent idea directory could be very valuable for prioritization, because in order to prioritize well, you would like to be able to brainstorm as many possible pros and cons for any given idea as possible.
To put it another way, getting more people involved in prioritization means we can take advantage of a broader array of perspectives. See Givewell on cluster thinking. I think having people brainstorm ways in which a proposal might end up being unexpectedly harmful could be especially valuable.
My impression is that idea production is like pottery in the sense that the best way to get quality is to aim for quantity.