Comment author: ThomasSittler 13 August 2017 08:29:56AM *  5 points [-]

Thanks for this impressive series of posts, Michael!

In the body text you write:

As far as I’m aware, there are no cost-effectiveness estimates comparing near-term causes like Give Directly and AMF to far-future ones, so I don’t know how much better X-risk charities are supposed to be.

But then you acknowledge in footnote 7 that there do exist such models. Michael Dickens' model is one example, and so are the models we developed during the Oxford Prioritisation Project, which you can tweak to your liking. Since not everyone may read the footnote, one thing you might do would be to amend the body text.

In a Part 2, you outline Six Ways DPR Could Do Good, yet your quantitative model only seems to take into account the first of these, improved treatment for mental health. Do you think it's the biggest effect? Or was it the easiest to model? Do you have plans to model other ways DPR could do good?

Assume the research caused by the rescheduling reveals ways to increase the happiness of each of these 10m people by 0.1 for a single year.

Does this assume that every single person suffering from depression or anxiety in the UK will adopt the new treatment? That seems unlikely. One way to get at this might be to look at current adoption rates of existing medication like SSRIs. These may be lower than adoption rates for new drugs would be, because SSRIs are less effective, but should give an interesting base rate.

You assume a 0.1 change. How does this relate to the studies on psilocybin you cite in part 2? For instance, you might look at how to translate changes on the QIDS (Quick Inventory of Depressive Symptomatology) scale into changes in HALYs. We did something very similar in our model of StrongMinds. In particular Konstantin and Lovisa wrote:

StrongMinds measures impact on the 27-point, linear PHQ-9 scale. To convert PHQ-9 impact to DALYs averted, Global Burden of Disease DALY-weighting of most severe depression (0.658) was divided by PHQ-9 points-weighting of most severe depression (27) to render 0.024 DALYs averted per PHQ-9 point reduced.

Finally, with regard to the £166m figure. It would be useful to get an order of magnitude estimate of the cost of past campaigns to achieve this kind of policy change, and their rate of success. A quick google scholar search throws up: Advancing the Empirical Research on Lobbying, Figueiredo and Richter, a literature review.

I want to stress that although my comment focused on areas of possible improvement in your post, I strongly support your doing this work and I think it's good. Good luck!

Comment author: DominikPeters 03 August 2017 08:11:23PM *  3 points [-]

Under certain circumstances, having moral uncertainty over theories that are purely ordinal may lead to the recommendation to split. Example: Suppose there are three charities A,B,C, and four options: donating 100% to one of A, B, C, or splitting the money equally between them (which we will call S). Let's ignore other ways of splitting. Suppose you have equal credence of 33% in three different theories:

1: A > S > B > C

2: B > S > C > A

3: C > S > A > B

Given the ranking over charities, it is rational in something like a von Neumann-Morgenstern sense to rank S second. But with these theories and these credences, one can see that S is the Condorcet winner and it is also the unique Borda winner, so that S would be uniquely recommended by essentially all voting rules, including Borda, the system favoured by Will MacAskill. In this example, contrary to the example in the OP, option S is not Pareto-dominated by another option, so that the unanimity principle does not bite.

This example crucially depends on only having ordinal information available, since with cardinal information (and expected value maximisation) we would never uniquely recommend splitting, as Tom notes, and so I don't think the argument in favour of splitting from moral uncertainty is particularly strong or robust.

Comment author: ThomasSittler 09 August 2017 11:16:56AM *  0 points [-]

Thanks, your counter-example is correct. I'm wondering how much of the effect is due to the way it individuates options? If we considered all possible ways of splitting S_1, S_2, ..., S_N, could we still construct sensible-seeming moral theories such that all (most?) voting rules recommend a significant amount of splitting?

In response to Introducing Enthea
Comment author: ThomasSittler 09 August 2017 11:11:56AM 2 points [-]

Interesting. Why focus on psilocybin and not other psychedelic substances? From reading the abstract, the systematic review you cite seems to claim similar effects for ayahuasca, psilocybin, and LSD.


How much donation splitting is there, and should you split?

Cross-posted from my blog . Table of contents Table of contents Summary Many of us split Individual examples EA funds data Naive approach: distribution of allocation percentages Less naive apprach: weighted distribution of allocation percentages Best approach: user totals Other data Arguments for splitting Empirical uncertainty combined with risk aversion... Read More
Comment author: ThomasSittler 28 July 2017 12:54:48PM 0 points [-]

The research notes file link is dead.

Comment author: SamDeere 04 July 2017 03:26:18AM *  3 points [-]

Short-medium term: some minor UI changes, to bring branding more into line with the rest of

Longer term ideas (caveat — these are just at the thought bubble stage at the moment and it's not clear whether they'd be valuable changes):

  • I think there's appetite for a discussion space that's both content aggregation as well as original content. This might take the form of getting a more active subreddit (for example) happening, but plausibly this could be something specifically built-for-purpose that either integrates with or complements the existing forum.

  • We've thought about integrating logins between the webapp on (what is currently just EA Funds) and the forum to avoid the need to manage multiple accounts when doing various EA things online

  • We've also thought a bit about integrating commenting systems so that discussion that happens on various EA blogs is mirrored on the forum (to avoid splitting discussions when cross-posting).

If there are things that you think would be useful (especially if you've been able to give this more thought than I have) that'd be great to know, with the caveat that we're pretty restricted by developer time on this, and the priority is ensuring ongoing maintenance of the existing infrastructure, rather than building out new features.

[eta spaces between dot points]

Comment author: ThomasSittler 04 July 2017 01:40:52PM *  2 points [-]

Other changes you might consider:

(1) Creating more than one category for posts (e.g. research, outreach, announcements, chat).

A first advantage is you can group by topic. The second big advantage I see from this is that different norms can develop for different categories. The threshold for posting in the EA forum is currently (perceived to be) too high for drafts or off-the-cuff ideas. So these either don't happen or move to Facebook (where discoverability is much worse). On the other hand, the prestige of posting, and the quality of discussion, is too low, so the most influential and busy EAs may not see it as a good use of their time. A way to improve this may be to have a very strictly moderated high-prestige section, and a more loosely moderated low-prestige section.

(2) Allowing markdown in the post composer

Comment author: Austen_Forrester 19 June 2017 10:06:38PM 1 point [-]

Have you considered combining the "GiveWell for impact investing" idea with the Effective Altruism Funds idea and create an EA impact investing biz within your charity? You could hire staff to find the best impact investing opportunities and create a few funds for different risk tolerances. Theoretically, it could pay for itself (or make serious money for CEA if successful enough) with a modest management fee. I'm not sure if charities are allowed to grant to businesses, but I know they can operate their own businesses as long as it's related to their mission.

Comment author: ThomasSittler 25 June 2017 08:35:23PM 0 points [-]

Hi Austen :) You may find this page useful, especially the "further reading".


QALYs/$ are more intuitive than $/QALYs

Cross-posted from my blog . Summary Cost-effectiveness estimates are often expressed in $/QALYs instead of QALYs/$. But QALYs/$ are preferable because they are more intuitive. To avoid small numbers, we can renormalise to QALYs/$10,000, or something similar. Cost-effectiveness estimates are often expressed in $/QALYs. Four examples: GiveWell, “Errors in DCP2... Read More
Comment author: ThomasSittler 31 May 2017 11:59:15PM 4 points [-]

It seems to me that the main reason people are sometimes insufficiently friendly or reliable is not the lack of papers detailing the benefits of considerateness.

Comment author: BenHoffman 21 May 2017 11:29:59PM 4 points [-]

Our prior strongly punishes MIRI. While the mean of its evidence distribution is 2,053,690,000 HEWALYs/$10,000, the posterior mean is only 180.8 HEWALYs/$10,000. If we set the prior scale parameter to larger than about 1.09, the posterior estimate for MIRI is greater than 1038 HEWALYs/$10,000, thus beating 80,000 Hours.

This suggests that it might be good in the long run to have a process that learns what prior is appropriate, e.g. by going back and seeing what prior would have best predicted previous years' impact.

Comment author: ThomasSittler 23 May 2017 11:18:47AM 2 points [-]

With the possible exception of StrongMinds, it's not the case that the previous years' impact is much easier to estimate than 2017's impact.

View more: Next