I’m interested in a collection of backchaining posts by EA organizations and individuals, that traces back from what we want—an optimal, safe, world—back to specific actions that individuals and groups can take.
Can be any level of granularity, though the more precise, the better.
Interested in this for any of the following categories:
I think a sort-of relevant collection can be found in the answers to this question about theory of change diagrams. And those answers also include other relevant discussion, like the pros and cons of trying to create and diagrammatically represent explicit theories of change. (A theory of change diagram won’t necessarily exactly meet your criteria, in the sense that it may backchain from an instrumental rather than intrinsic goal, but it’s sort-of close.)
I think Leverage’s diagram might be the closest thing I know of to a fairly granular backchaining from one’s ultimate goals. It also seems to me quite unwieldy—I spent a while trying to read it once, but it felt annoying to navigate and hard to really get the overall gist of. (That was just my personal take, though.)
One could also argue that Toby Ord’s “grand strategy for humanity”[1] is a very low-granularity instance of backchaining from one’s ultimate goals. And it becomes more granular once one connects the first step of the grand strategy to other specific recommendations Ord makes in The Precipice.
(I know you and I have already discussed some of this; this comment was partly for other potential readers’ sake.)
[1] For readers who haven’t read The Precipice, Ord’s quick summary of the grand strategy is as follows:
I think that at the highest level we should adopt a strategy proceeding in three phases:
Reaching Existential Security
The Long Reflection
Achieving Our Potential
The book contains many more details on these terms and this strategy, of course.
I’m interested in a collection of backchaining posts by EA organizations and individuals, that traces back from what we want—an optimal, safe, world—back to specific actions that individuals and groups can take.
Can be any level of granularity, though the more precise, the better.
Interested in this for any of the following categories:
Effective Altruism
Longtermism
General GCR reduction
AI Safety
Biorisk
Institutional decision-making
I think a sort-of relevant collection can be found in the answers to this question about theory of change diagrams. And those answers also include other relevant discussion, like the pros and cons of trying to create and diagrammatically represent explicit theories of change. (A theory of change diagram won’t necessarily exactly meet your criteria, in the sense that it may backchain from an instrumental rather than intrinsic goal, but it’s sort-of close.)
The answers in that post include links to theory of change diagrams from Animal Charity Evaluators (p.15), Convergence Analysis, Happier Lives Institute, Leverage Research, MIRI, and Rethink Priorities. Those are the only 6 research orgs I know of which have theory of change diagrams. (But that question was just about research orgs, and having such diagrams might be somewhat more common among non-research organisations.)
I think Leverage’s diagram might be the closest thing I know of to a fairly granular backchaining from one’s ultimate goals. It also seems to me quite unwieldy—I spent a while trying to read it once, but it felt annoying to navigate and hard to really get the overall gist of. (That was just my personal take, though.)
One could also argue that Toby Ord’s “grand strategy for humanity”[1] is a very low-granularity instance of backchaining from one’s ultimate goals. And it becomes more granular once one connects the first step of the grand strategy to other specific recommendations Ord makes in The Precipice.
(I know you and I have already discussed some of this; this comment was partly for other potential readers’ sake.)
[1] For readers who haven’t read The Precipice, Ord’s quick summary of the grand strategy is as follows:
The book contains many more details on these terms and this strategy, of course.
It has occurred to me that very few such documents exist.
I’m curious what it looks like to backchain from something so complex. I’ve tried it repeatedly in the past and feel like I failed.