Hi David. Thanks for the post.

I have an annual reminder to tell my closest family about the charities I would like to leave 90 % of my assets to if I die. I tell them they could take the other 10 %, which I think results in a greater fraction going to charity than if I told them to leave 100 % to charity.

I considered setting up a will, but I concluded it is not worth it for me. I estimate it would cost 234 € for financial cost of 159 € in Portugal, and time cost of 75 € for 5 h at 15 €/h. The risk of death of people in Portugal in 2023 with age 25 to 29, which covers my age, was 0.0416 %. So it would only be worth it if I could lose more than 562 k€ (= 234/(4.16*10^-4)). My total assets are much smaller than this, and I expect this to continue to be the case. I aim to keep my savings equal to 6 times the global real gross domestic product (GDP) per capita, which is currently around 80 k€ in Portugal.

I replied here.

Do you think pain A having a higher intensity than B implies that averting an infinitesimal duration of pain A with an infinitesimal probability is better than averting an astronomically long time of pain B with certainty? If not, what is required for you to believe this besides A having a higher pain intensity than B?

Consider someone holding their hand in hot water for 1 min. If you think there are only 5 pain intensities, what would be the range of temperature for each pain intensity? If the ith pain intensity starts at T_(i - 1), and ends at T_i, what would change so much from temperature T_i_before = T_i - 0.001 ºC to T_i_after = T_i + 0.001 ºC that makes you prioritise averting pain at T_i_after infinitely more than pain at T_i_before? Do you believe empirical studies of people's preferences would find a few temperatures (4 if you believe in 5 pain intensities) with this property, where people would prefer averting any time in pain at temperature T_i_after over an arbitrarily long time in pain at temperature T_i_before?

Hi Vince.

ACE proudly recommends Sociedade Vegetariana Brasileira as an excellent giving opportunity. To learn more, read our 2025 comprehensive review of Sociedade Vegetariana Brasileira.

This just links to their logo. Here is ACE's review.

I suspect you would prefer averting A for any value of N (equal to 1 or higher) in principle, but that you believe N cannot take a high value in practice.

I understand I got this right. So, if N could be 1 M, I think you would prefer averting i) 1 year of pain of intensity level 1 M with probability 10^-100 over ii) 10^100 years of pain of intensity level 999,999 with probability 1.

While mathematical models can theoretically increase N to infinity, the subjective reality of suffering is biologically capped because there is a physiological ceiling of the nervous system. 

N is supposed to be the number of different pain intensities. One cannot determine the maximum pain intensity M based on N alone. In theory, N can be arbitrarily large for any M. The mean difference Delta = M/N between consecutive pain intensity levels would just tend to 0 as N increases to infinity. For a constant difference between the pain intensity of consecutive intensity levels, the pain intensity of level i would be Delta*i.

The premise of "99.99999999% of X" assumes that pain exists on a perfectly smooth, linear scale that can be infinitely divided.

I agree pain intensities cannot be arbitrarily close. However, consider N = 100. Would you prefer averting, for example, i) 0.1 s of pain of intensity level 100 with probability 10^-100 over ii) 10^100 years of pain of intensity level 99 with probability 1. The expected pain of ii) is 3.12*10^208 (= 10^100*99*1/(0.1/60^2/24/365.25*100*10^-100)) times that of i).

It also seems something worth tracking to me.

Thanks for the post, Lauren. Here is a discussion about the post.

Here is Lauren's post about this.

Thanks for elaborating. Imagine pain has N levels of intensity (with N equal to 1 or higher). Consider the following pains:

• A. Duration of 1 year, intensity level N (the highest), and probability of 10^-100.
• B. Duration of 10^100 years, intensity level N - 1 (the 2nd highest), and probability of 1.

The expected pain of B is 10^100*(N - 1)*1/(1*N*10^-100) = 10^200*(1 - 1/N) times that of A. I would prefer averting B for N equal to 2 or higher. Even for N = 2, the expected pain of B is 5*10^199 times that of A.

For which values of N (if any) would you prefer averting B over A? I understand you would prefer averting A at least for N = 5. I suspect you would prefer averting A for any value of N (equal to 1 or higher) in principle, but that you believe N cannot take a high value in practice. If so, why?

Hi David. I would be curious to know your thoughts on my reply to titotal. In the post, "more pain" is supposed to mean "more pain, and all else equal". If the event that leads to additional pain comes with benefits, it could overall increase welfare. In contrast, a "speck of dust in the eye" is supposed to represent something which decreases welfare very little (considering all effects).