Are you only talking about donations? Do you think it would also be a mistake to work on speculative causes? (That seems much different in that there are many more opportunities for learning from working on a cause than from donating to it.)
I agree with that point. I’m not talking only about donations per se, though that’s a much more important consideration for the “typical person”. I think in so far as one should be purchasing either impact through proven causes or value of information through speculative causes, value of information is better purchased by work rather than money.
~
I think it can make sense for other people with the right kind of information to fund other opportunities. E.g., think about early GiveWell.
I agree. This is why I want to see certain “promising” causes funded, provided we will get to see if they succeed or fail.
~
What counts as a “speculative cause”? Meta-research? Political advocacy? Education for talented youth? Climate change? Prioritization work? Funding early GiveWell? Funding 80K? Anyone who says the thing they are doing is somehow improving very long-run outcomes? Anything that hasn’t been recommended by GiveWell? Anything that hasn’t been proven to work with RCT-quality evidence?
I’d say a “speculative cause” is any cause who’s impact has not been demonstrated with empirical evidence of sufficient quality. It doesn’t have to be an RCT. For example, vegetarianism advocacy could demonstrate impact, in my opinion, with just a few studies with actual control groups run by independent people that still show the conversion rate of ~2%.
I think Giving What We Can or 80K could sufficiently demonstrate their impact with just slightly better member surveys and slightly more robust analyses.
For another example, I think GiveWell itself has proven itself to be a good cause (albeit not one with more funding) through their documentation of money moved and track record of good research and there aren’t any RCTs done on GiveWell.
~
If what you care about is the far future, why does AMF get to count as a “non-speculative” giving opportunity? We have very little idea how much AMF improves very long-run outcomes in comparison with other things, little effort has gone into studying this, and it seems many of your arguments that we are bad at understanding long-run effects apply.
I think this is a key thing to clear up. AMF is non-speculative in terms of producing near-term impact (or impact generically), but it is still speculative in terms of long-term impact. I think if you want to purchase direct impact (instead of information), you should be doing so through a proven cause like AMF.
Your last point doesn’t make much sense to me. I agree that we should be concerned about purchasing as much impact as we can, but the amount of impact you’re purchasing from AMF is minuscule compared to the far future. It seems like your concern for something being ‘proven’ is skewing your choices.
It’s like walking into a shop and buying the first TV (or whatever) that you see, despite it likely being expensive and not nearly as good as other ones, because you can at least see it, rather than doing a bit of looking on Amazon.
If you look for TVs among Amazon, they give you specific prices and you can reliably buy them. Now imagine you walked into a TV shop and saw some that were a good deal (say $100), but others where the price was set randomly?
I don’t think we can make a case that we’re actually purchasing some known, larger-than-AMF amount of impact from far future work, because so much far future work is likely to not actually do well at affecting the far future. I just don’t think we have that much control over how the future unfolds.
Some interventions have no impact, some have low impact, and some have high impact. ‘No impact’ doesn’t help anyone/do any good, ‘low impact’ helps some people/does some good, and ‘high impact’ helps a lot of people/does a lot of good. Because of the size of the future, an intervention has to help a lot of people/do a lot of good to be ‘high-impact’ - helping millions or billions rather than thousands or tens of thousands.
We’re fairly sure that AMF is ‘low impact’, since we have evidence that it reliably helps a decent number of present people. Which is great—it’s not ‘no impact’! But it’s unlikely that it will be ‘high impact’.
I agree that we don’t have a clear sense yet of which interventions are actually high-impact. That’s why I don’t donate to any direct x-risk reduction effort. However the appropriate response to this problem seems to be to invest in more research, to work out which interventions will plausibly be high-impact. Alternatively, one could invest in improving one’s position to be able to purchase more of the high-impact intervention when we have a clearer view of what that is—putting oneself one a good career path or building an effective movement.
I don’t understand why you think the response should be to purchase low-impact interventions.
I agree that we don’t have a clear sense yet of which interventions are actually high-impact. [...] However the appropriate response to this problem seems to be to invest in more research, to work out which interventions will plausibly be high-impact. Alternatively, one could invest in improving one’s position to be able to purchase more of the high-impact intervention when we have a clearer view of what that is—putting oneself one a good career path or building an effective movement.
I definitely agree with this, and that’s what I tried to articulate in the section on value of information. Sorry if I wasn’t clear.
About donating vs. working: I don’t agree that the two are fundamentally different. Basically what I’m saying is time is money and money is time. I think there is a conversion rate between the two and that rate is certainly not static. But you have to consider the counterfactual, what you could be doing if you didn’t spend your time or money on a cause. For example if you are a very good cost-effectiveness researcher your time may be worth 2 or 3 times as much as the money you could make earning to give. But that money could pay for at least one additional researcher. Similarly, if you spend your time writing web posts that time could be spent working a student job or better, investing in your career which will yield money in the long run. Somehow you can always convert the two.
Although we can’t quantify the conversion rate I think it exists. And generally I would expect it to be around 1, maybe between 0,33 and 3 in some cases. That is sometimes significant, but with the different causes we’re talking about we don’t expect the impact to be even in the same order of magnitude. So for example it wouldn’t make sense to say working on MIRI is a good idea, but donating to them is not a good idea.
So if I say I would not donate to an organisation it makes sense to ask if I would not work for that organisation either. Money and time are somehow convertible into each other. So I wouldn’t agree that value of information is always better purchased by time than by money. And if it is, the difference may not be all that great.
And a question: Maybe this has been answered somewhere, but I think it would be useful if you straightforwardly said which causes you consider speculative. Makes this quite vague discussion a little less vague hopefully. So say 80k, EAA, GWWC, MIRI, FHI, CFAR, Effective Fundraising, Animal Ethics, all x-risk related, all research related?? This would really help me. I think the issue you raised with your post is very important to discuss and get right ;)
So if I say I would not donate to an organisation it makes sense to ask if I would not work for that organisation either.
I’m not sure this is the case. If you’re working for the organiazation, you’re in a significant different place with regard to the amount of information you can get on the organization’s impact and what you can do to increase that information. I think it’s possible to be the case that if you’re working in a speculative organization, you can get further information through working but not through donating.
~
Maybe this has been answered somewhere, but I think it would be useful if you straightforwardly said which causes you consider speculative. Makes this quite vague discussion a little less vague hopefully
I see, that makes sense. I understood value of information as creating valuable information for the whole community. You seem to be talking about valuable information for oneself. And maybe as an added bonus increasing the information about your organization more than a replacement worker would otherwise.
But yes, it makes sense to me that if you work for a speculative cause you are in a better position to assess if you should donate to them.
The point I was trying to make is less about value of information for yourself but information for others. Your donation ould fund a new employee for example who 1. gathers a lot of information like you would if you were in there position and 2. brings valuable information to the community in general. The questions is of course whether that person would be as productive as you.
I understood value of information as creating valuable information for the whole community. You seem to be talking about valuable information for oneself.
Well, any information I gather individually could be shared.
~
Your donation could fund a new employee for example who 1. gathers a lot of information like you would if you were in there position and 2. brings valuable information to the community in general.
Right. That would be one way to do it, if you could trust the person you hire to be interested in gathering information. Right now, my perception is that people who are interested in gathering information and reporting it are kind of rare.
That’s an interesting point. I strongly agree that less proven charities should do more internal research and especially reporting about their effectiveness. I think this could fuel an important discussion. Even if the results aren’t that amazing I think certain people would consider donating to them more simply because they are more aware of the opportunity and/or less uncertain about it.
I’m not quite sure yet what exactly you refer to by information though. It sounds like this refers to reporting about the effectiveness of the charities. Or are you talking about information like cost-effectiveness research and research papers/blog posts as well?
I agree with that point. I’m not talking only about donations per se, though that’s a much more important consideration for the “typical person”. I think in so far as one should be purchasing either impact through proven causes or value of information through speculative causes, value of information is better purchased by work rather than money.
~
I agree. This is why I want to see certain “promising” causes funded, provided we will get to see if they succeed or fail.
~
I’d say a “speculative cause” is any cause who’s impact has not been demonstrated with empirical evidence of sufficient quality. It doesn’t have to be an RCT. For example, vegetarianism advocacy could demonstrate impact, in my opinion, with just a few studies with actual control groups run by independent people that still show the conversion rate of ~2%.
I think Giving What We Can or 80K could sufficiently demonstrate their impact with just slightly better member surveys and slightly more robust analyses.
For another example, I think GiveWell itself has proven itself to be a good cause (albeit not one with more funding) through their documentation of money moved and track record of good research and there aren’t any RCTs done on GiveWell.
~
I think this is a key thing to clear up. AMF is non-speculative in terms of producing near-term impact (or impact generically), but it is still speculative in terms of long-term impact. I think if you want to purchase direct impact (instead of information), you should be doing so through a proven cause like AMF.
Your last point doesn’t make much sense to me. I agree that we should be concerned about purchasing as much impact as we can, but the amount of impact you’re purchasing from AMF is minuscule compared to the far future. It seems like your concern for something being ‘proven’ is skewing your choices.
It’s like walking into a shop and buying the first TV (or whatever) that you see, despite it likely being expensive and not nearly as good as other ones, because you can at least see it, rather than doing a bit of looking on Amazon.
I’d reject your analogy.
If you look for TVs among Amazon, they give you specific prices and you can reliably buy them. Now imagine you walked into a TV shop and saw some that were a good deal (say $100), but others where the price was set randomly?
I don’t think we can make a case that we’re actually purchasing some known, larger-than-AMF amount of impact from far future work, because so much far future work is likely to not actually do well at affecting the far future. I just don’t think we have that much control over how the future unfolds.
Some interventions have no impact, some have low impact, and some have high impact. ‘No impact’ doesn’t help anyone/do any good, ‘low impact’ helps some people/does some good, and ‘high impact’ helps a lot of people/does a lot of good. Because of the size of the future, an intervention has to help a lot of people/do a lot of good to be ‘high-impact’ - helping millions or billions rather than thousands or tens of thousands.
We’re fairly sure that AMF is ‘low impact’, since we have evidence that it reliably helps a decent number of present people. Which is great—it’s not ‘no impact’! But it’s unlikely that it will be ‘high impact’.
I agree that we don’t have a clear sense yet of which interventions are actually high-impact. That’s why I don’t donate to any direct x-risk reduction effort. However the appropriate response to this problem seems to be to invest in more research, to work out which interventions will plausibly be high-impact. Alternatively, one could invest in improving one’s position to be able to purchase more of the high-impact intervention when we have a clearer view of what that is—putting oneself one a good career path or building an effective movement.
I don’t understand why you think the response should be to purchase low-impact interventions.
I definitely agree with this, and that’s what I tried to articulate in the section on value of information. Sorry if I wasn’t clear.
About donating vs. working: I don’t agree that the two are fundamentally different. Basically what I’m saying is time is money and money is time. I think there is a conversion rate between the two and that rate is certainly not static. But you have to consider the counterfactual, what you could be doing if you didn’t spend your time or money on a cause. For example if you are a very good cost-effectiveness researcher your time may be worth 2 or 3 times as much as the money you could make earning to give. But that money could pay for at least one additional researcher. Similarly, if you spend your time writing web posts that time could be spent working a student job or better, investing in your career which will yield money in the long run. Somehow you can always convert the two.
Although we can’t quantify the conversion rate I think it exists. And generally I would expect it to be around 1, maybe between 0,33 and 3 in some cases. That is sometimes significant, but with the different causes we’re talking about we don’t expect the impact to be even in the same order of magnitude. So for example it wouldn’t make sense to say working on MIRI is a good idea, but donating to them is not a good idea.
So if I say I would not donate to an organisation it makes sense to ask if I would not work for that organisation either. Money and time are somehow convertible into each other. So I wouldn’t agree that value of information is always better purchased by time than by money. And if it is, the difference may not be all that great.
And a question: Maybe this has been answered somewhere, but I think it would be useful if you straightforwardly said which causes you consider speculative. Makes this quite vague discussion a little less vague hopefully. So say 80k, EAA, GWWC, MIRI, FHI, CFAR, Effective Fundraising, Animal Ethics, all x-risk related, all research related?? This would really help me. I think the issue you raised with your post is very important to discuss and get right ;)
I’m not sure this is the case. If you’re working for the organiazation, you’re in a significant different place with regard to the amount of information you can get on the organization’s impact and what you can do to increase that information. I think it’s possible to be the case that if you’re working in a speculative organization, you can get further information through working but not through donating.
~
I answer this now in “What Would it Take to ‘Prove’ A Speculative Cause?”.
I see, that makes sense. I understood value of information as creating valuable information for the whole community. You seem to be talking about valuable information for oneself. And maybe as an added bonus increasing the information about your organization more than a replacement worker would otherwise.
But yes, it makes sense to me that if you work for a speculative cause you are in a better position to assess if you should donate to them.
The point I was trying to make is less about value of information for yourself but information for others. Your donation ould fund a new employee for example who 1. gathers a lot of information like you would if you were in there position and 2. brings valuable information to the community in general. The questions is of course whether that person would be as productive as you.
Well, any information I gather individually could be shared.
~
Right. That would be one way to do it, if you could trust the person you hire to be interested in gathering information. Right now, my perception is that people who are interested in gathering information and reporting it are kind of rare.
That’s an interesting point. I strongly agree that less proven charities should do more internal research and especially reporting about their effectiveness. I think this could fuel an important discussion. Even if the results aren’t that amazing I think certain people would consider donating to them more simply because they are more aware of the opportunity and/or less uncertain about it.
I’m not quite sure yet what exactly you refer to by information though. It sounds like this refers to reporting about the effectiveness of the charities. Or are you talking about information like cost-effectiveness research and research papers/blog posts as well?
I’m thinking here information about impact, or evidence that would lower our uncertainty about the effect of a certain intervention.