Staying in contact with the incredibly-serious, but not literally-most-important problems in the world seems healthy for personal calibration and credibly conveying the values that drive your work to people who might otherwise not understand.
Following expected value, but being less than a complete slave to it seems like the right long run strategy from most perspectives. I think this is your former employer's implied policy for good reason!
> In fact, the very next day after we made this donation, I found myself in a conversation full of longtermist EAs worrying that all the newly minted billionaires would quickly spend down their money on global health and animal well-being interventions instead of saving it for more removed but ultimately higher-leverage opportunities that crop up during (or even after) the Singularity.
I'm sorry, and I appreciate the post so apologies for being unkind, but I just think this loses the plot a little bit on some of their parts.
I think you should follow the math *extremely* far, but I also think of the idea of reflective salience; that the issues you spend a lot of time discussing are the issues you end up giving the most attention, and it just seems like "Oh no, these people making extremely net-positive contributions could have made even more extremely net-positive contributions" is a deep misallocation of your mental energy when there exist so many people close to zero or at net-negative you could be working to try to convince to move somewhat.
>Oh no, these people making extremely net-positive contributions could have made even more extremely net-positive contributions" is a deep misallocation of your mental energy when there exist so many people close to zero or at net-negative you could be working to try to convince to move somewhat.
I think this is a reasonable heuristic for most people at most times, but it’s not correct in this case. The people in that room are in a position to make arguments to substantially influence a staggering amount of philanthropic capital from people who share our broad perspectives and are very actively deciding where to give; we would not in fact have more impact by thinking about how to appeal to people who are less bought in on the basic ideas. (To be clear, we all work in jobs other than philanthropic capital allocation most of the time, but our philanthropic capital allocation discussions should in fact be focused on this particular group of people imo.)
I think the indirect benefit of such a donation to Ai safety could be quite substantial. It is common to when trying to win influence and drive change to seek visible quick wins, that then allow longer term work to be done. Animal welfare and the classic malaria type work can be thought of as such quick wins, especially if tangible results can be shown
Separately you are allowed to do what makes you happy - that also keeps you in the game
And then there's also the somewhat abstract argument that how we treat animals may influence how an ASI treats us
Staying in contact with the incredibly-serious, but not literally-most-important problems in the world seems healthy for personal calibration and credibly conveying the values that drive your work to people who might otherwise not understand.
Following expected value, but being less than a complete slave to it seems like the right long run strategy from most perspectives. I think this is your former employer's implied policy for good reason!
> In fact, the very next day after we made this donation, I found myself in a conversation full of longtermist EAs worrying that all the newly minted billionaires would quickly spend down their money on global health and animal well-being interventions instead of saving it for more removed but ultimately higher-leverage opportunities that crop up during (or even after) the Singularity.
I'm sorry, and I appreciate the post so apologies for being unkind, but I just think this loses the plot a little bit on some of their parts.
I think you should follow the math *extremely* far, but I also think of the idea of reflective salience; that the issues you spend a lot of time discussing are the issues you end up giving the most attention, and it just seems like "Oh no, these people making extremely net-positive contributions could have made even more extremely net-positive contributions" is a deep misallocation of your mental energy when there exist so many people close to zero or at net-negative you could be working to try to convince to move somewhat.
>Oh no, these people making extremely net-positive contributions could have made even more extremely net-positive contributions" is a deep misallocation of your mental energy when there exist so many people close to zero or at net-negative you could be working to try to convince to move somewhat.
I think this is a reasonable heuristic for most people at most times, but it’s not correct in this case. The people in that room are in a position to make arguments to substantially influence a staggering amount of philanthropic capital from people who share our broad perspectives and are very actively deciding where to give; we would not in fact have more impact by thinking about how to appeal to people who are less bought in on the basic ideas. (To be clear, we all work in jobs other than philanthropic capital allocation most of the time, but our philanthropic capital allocation discussions should in fact be focused on this particular group of people imo.)
I think the indirect benefit of such a donation to Ai safety could be quite substantial. It is common to when trying to win influence and drive change to seek visible quick wins, that then allow longer term work to be done. Animal welfare and the classic malaria type work can be thought of as such quick wins, especially if tangible results can be shown
Separately you are allowed to do what makes you happy - that also keeps you in the game
And then there's also the somewhat abstract argument that how we treat animals may influence how an ASI treats us
Thank you for posting about this. I had no idea, and I'm now shaken to the core