EDIT: This is now superseded by a top-level post so you should read that instead.
tl;dr: Value large impacts rather than large inputs, but be excited about megaprojects anyway because they’re a new & useful tool we’ve unlocked
A lot of people are excited about megaprojects, and I agree that they should be. But we should remember that megaprojects are basically defined by the size of their inputs (e.g., “productively” using >$100 million per year), and that we don’t intrinsically value the capacity to absorb those inputs. What we really care about is huge positive impact, and megaprojects are just one means to that end, and actually (ceteris paribus) we should be even more excited about achieving the same impacts using less inputs & smaller projects. How can we reconcile these thoughts, and why should we still be excited about megaprojects?
I suggest we think about this as follows:
Imagine a Venn diagram with a circle for megaprojects and another circle for projects with great expected value (EV)
Projects with great EV are really the focus and always have been
Projects like 80,000 Hours, FHI, and Superintelligence were each far smaller than megaprojects, but in my view probably had and still have great EV, and an EV high enough to potentially justify megaproject-level spending. That’s great (even better than megaprojects!), and we’d still love more projects that can punch so far above their weight.
But there’s also a large area of overlap between the two circles of the Venn diagram—a large overlap between megaprojects and projects with great EV—since it will usually take a lot of inputs to achieve a lot of good outcomes.
And we haven’t yet explored that area of overlap much—haven’t found and executed on the obvious and best ideas. This is partly because it obviously generally makes sense to start with the smaller, cheaper, easier options, and partly because EA’s stock of financial resources and relevant human capital has grown fairly rapidly so just a few years ago we had far less ability to execute on megaprojects. It’s probably also partly because a lot of people aren’t naturally sufficiently ambitious or lack sufficient self-confidence.
Now that those stocks of financial and human capital resources have grown so much, we’ve sort-of “unlocked” that additional area of high-EV project options. And we’re likely continuing to unlock it more each year.
So we should now be explicitly focusing attention on that area of options; if we don’t want make an explicit effort to do that, we’ll continue neglecting it via inertia. This doesn’t make smaller projects that also have great EV any less valuable, but we don’t want to be only focusing on those.
But we should remember that really we should first and foremost be extremely ambitious in terms of impacts, and just willing to also—as a means to that end—be extremely ambitious in terms of inputs absorbed.
One caveat: We should also to some extent value doing larger projects for the sake of doing larger projects—like sometimes choose that over doing smaller projects with similar/great impact—since that upgrades the career capital of the project’s leaders/employees and also provides community-level lessons learned, helping further unlock the option of future megaprojects. But this still isn’t a matter of valuing the size of projects as an end in itself.
Caveats: I wrote this fairly quickly and didn’t run it by people to check that it clearly conveys my full views here. I imagine some people could take this as “be less excited about megaprojects”, which is definitely not a message I want to convey to EA in general.
My thanks to Linch Zhang for conversations that informed my thinking here, though that doesn’t imply his endorsement of my thinking or of this shortform.
I think the general thrust of your argument is clearly right, and it’s weird/frustrating that this is not the default assumption when people talk about megaprojects (though maybe I’m not reading the existing discussions of megaprojects sufficiently charitably).
2 moderately-sized caveats:
Re 2) “Projects with great EV are really the focus and always have been”, I think in the early days of EA, and to a lesser degree still today, a lot of focus of EA isn’t on great EV so much as high cost-effectiveness. To some degree the megaprojects discourse was set to push back against this.
Re: 5, “It’s probably also partly because a lot of people aren’t naturally sufficiently ambitious or lack sufficient self-confidence” I think this is definitely true, but maybe I’d like to push back a bit on the individual framing of this lack of ambition, as I think it’s partially cultural/institutional. That is, until very recently, we (EA broadly, or the largest funders etc), haven’t made it as clear that EA supports and encourages extreme ambition in outputs in a way that means we (collectively) are potentially willing to pay large per-project costs in inputs.
Thanks—I think those are both really good points! I’ve now made a top-level post version of this shortform, with the main modifications being adjustments in light of your points (plus, unrelatedly, adding a colourful diagram because colourful diagrams are fun).
EDIT: This is now superseded by a top-level post so you should read that instead.
tl;dr: Value large impacts rather than large inputs, but be excited about megaprojects anyway because they’re a new & useful tool we’ve unlocked
A lot of people are excited about megaprojects, and I agree that they should be. But we should remember that megaprojects are basically defined by the size of their inputs (e.g., “productively” using >$100 million per year), and that we don’t intrinsically value the capacity to absorb those inputs. What we really care about is huge positive impact, and megaprojects are just one means to that end, and actually (ceteris paribus) we should be even more excited about achieving the same impacts using less inputs & smaller projects. How can we reconcile these thoughts, and why should we still be excited about megaprojects?
I suggest we think about this as follows:
Imagine a Venn diagram with a circle for megaprojects and another circle for projects with great expected value (EV)
Projects with great EV are really the focus and always have been
Projects like 80,000 Hours, FHI, and Superintelligence were each far smaller than megaprojects, but in my view probably had and still have great EV, and an EV high enough to potentially justify megaproject-level spending. That’s great (even better than megaprojects!), and we’d still love more projects that can punch so far above their weight.
But there’s also a large area of overlap between the two circles of the Venn diagram—a large overlap between megaprojects and projects with great EV—since it will usually take a lot of inputs to achieve a lot of good outcomes.
And we haven’t yet explored that area of overlap much—haven’t found and executed on the obvious and best ideas. This is partly because it obviously generally makes sense to start with the smaller, cheaper, easier options, and partly because EA’s stock of financial resources and relevant human capital has grown fairly rapidly so just a few years ago we had far less ability to execute on megaprojects. It’s probably also partly because a lot of people aren’t naturally sufficiently ambitious or lack sufficient self-confidence.
Now that those stocks of financial and human capital resources have grown so much, we’ve sort-of “unlocked” that additional area of high-EV project options. And we’re likely continuing to unlock it more each year.
So we should now be explicitly focusing attention on that area of options; if we don’t want make an explicit effort to do that, we’ll continue neglecting it via inertia. This doesn’t make smaller projects that also have great EV any less valuable, but we don’t want to be only focusing on those.
But we should remember that really we should first and foremost be extremely ambitious in terms of impacts, and just willing to also—as a means to that end—be extremely ambitious in terms of inputs absorbed.
One caveat: We should also to some extent value doing larger projects for the sake of doing larger projects—like sometimes choose that over doing smaller projects with similar/great impact—since that upgrades the career capital of the project’s leaders/employees and also provides community-level lessons learned, helping further unlock the option of future megaprojects. But this still isn’t a matter of valuing the size of projects as an end in itself.
Caveats: I wrote this fairly quickly and didn’t run it by people to check that it clearly conveys my full views here. I imagine some people could take this as “be less excited about megaprojects”, which is definitely not a message I want to convey to EA in general.
My thanks to Linch Zhang for conversations that informed my thinking here, though that doesn’t imply his endorsement of my thinking or of this shortform.
I think the general thrust of your argument is clearly right, and it’s weird/frustrating that this is not the default assumption when people talk about megaprojects (though maybe I’m not reading the existing discussions of megaprojects sufficiently charitably).
2 moderately-sized caveats:
Re 2) “Projects with great EV are really the focus and always have been”, I think in the early days of EA, and to a lesser degree still today, a lot of focus of EA isn’t on great EV so much as high cost-effectiveness. To some degree the megaprojects discourse was set to push back against this.
Re: 5, “It’s probably also partly because a lot of people aren’t naturally sufficiently ambitious or lack sufficient self-confidence” I think this is definitely true, but maybe I’d like to push back a bit on the individual framing of this lack of ambition, as I think it’s partially cultural/institutional. That is, until very recently, we (EA broadly, or the largest funders etc), haven’t made it as clear that EA supports and encourages extreme ambition in outputs in a way that means we (collectively) are potentially willing to pay large per-project costs in inputs.
Thanks—I think those are both really good points! I’ve now made a top-level post version of this shortform, with the main modifications being adjustments in light of your points (plus, unrelatedly, adding a colourful diagram because colourful diagrams are fun).