Geoff_Anders comments on Leverage Research: reviewing the basic facts - Effective Altruism Forum

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (64)

You are viewing a single comment's thread.

Comment author: Geoff_Anders 03 August 2018 04:15:03PM *  25 points [-]

Hi everyone,

I'd like to start off by apologizing. I realize that it has been hard to understand what Leverage has been doing, and I think that that's my fault. Last month Kerry Vaughan convinced me that I needed a new approach to PR and public engagement, and so I've been thinking about what to write and say about this. My plan, apart from the post here, was to post something over the next month. So I'll give a brief response to the points here and then make a longer main post early next week.

(1) I'm sorry for the lack of transparency and public engagement. We did a lot more of this in 2011-2012, but did not really succeed in getting people to understand us. After that, I decided we should just focus on our research. I think people expect more public engagement, even very early in the research process, and that I did not understand this.

(2) We do not consider ourselves an EA organization. We do not solicit funds from individual EAs. Instead, we are EA-friendly, in that (a) we employ many EAs, (b) we let people run EA projects, and (c) we contribute to EA causes, especially EA movement building. As noted in the post, we ran the EA Summit 2013 and EA Summit 2014. These were the precursors to the EA Global conferences. For a sense of what these were like, see the EA Summit 2013 video. We also ran the EA Retreat 2014 and helped out operationally with EA Global 2015. We also founded THINK, the first EA movement group network.

(3) We are probably not the correct donation choice for most EAs. We care about reason, evidence, and impact, but we are much higher variance than most EAs would like. We believe there is evidence that we are more effective than regular charities due to our contributions to movement building. These can be thought of as "impact offsets". (See (6) for more on the EV calculation.)

(4) We are also probably not the correct employment choice for most EAs. We are looking for people with particular skills and characteristics (e.g., ambition, dedication to reason and evidence, self-improvement). These make CFAR our largest EA competitor for talent, though in actual fact we have not ended up competing substantially with them. In general if people are excited about CEA or 80k or Charity Science or GiveWell or OPP, then we typically also expect that they are better served by working there.

(5) Despite (3) and (4), we are of course very interested in meeting EAs who would be good potential donors or recruits. We definitely recruit at EA events, though again we think that most EAs would be better served by working elsewhere.

(6) To do a full EV calculation on Leverage, it is important to take into account the counterfactual cost of employees who would work on other EA projects. We think that taking this into account, counting our research as 0 value, and using the movement building impact estimates from LEAN, we come out well on EV compared to an average charity. This is because of our contribution to EA movement building and because EA movement building is so valuable. (Rather than give a specific Fermi estimate, I will let readers make their own calculations.) Of course, under these assumptions donating to movement building alone is higher EV than donating to Leverage. Donors should only consider us if they assign greater than 0 value to our research.

I hope that that clarifies to some degree Leverage's relation to the EA movement. I'll respond to the specific points above later today.

As for the EA Summit 2018, we agree that everyone should talk with people they know before attending. This is true of any multi-day event. Time is valuable, and it's a good idea to get evidence of the value prior to attending.

(Leverage will not be officially presenting any content at the EA Summit 2018, so people who would like to learn more should contact us here. My own talk will be about how to plan ambitious projects.)

EDIT: I said in my earlier comment that I would write again this evening. I’ll just add a few things to my original post.

— Many of the things listed in the original post are simply good practice. Workshops should track participants to ensure the quality of their experience and that they are receiving value. CFAR also does this. Organizations engaged in recruitment should seek to proactively identify qualified candidates. I’ve spoken to the leaders of multiple organizations who do this.

— Part of what we do is help people to understand themselves better via introspection and psychological frameworks. Many people find this both interesting and useful. All of the mind mapping we did was with the full knowledge and consent of the person, at their request, typically with them watching and error-checking as we went. (I say “did” because we stopped making full mind maps in 2015.) This is just a normal part of showing people what we do. It makes sense for prospective recruits and donors to seek an in-depth look at our tools prior to becoming more involved. We also have strict privacy rules and do not share personal information from charting sessions without explicit permission from the person. This is true for everyone we work with, including prospective recruits and donors.

Comment author: Jeff_Kaufman 04 August 2018 02:34:54PM 30 points [-]

Hi Geoff,

In reading this I'm confused about the relationship between Paradigm and Leverage. People in this thread (well, mostly Evan) seem to be talking about them as if Leverage incubated Paradigm but the two are now fully separate. My understanding, however, was that the two organizations function more like two branches of a single entity? I don't have a full picture or anything, but I thought you ran both organizations, staff of both mostly live at Leverage, people move freely between the two as needed by projects, and what happens under each organization is more a matter of strategy than separate direction?

By analogy, I had thought the relationship of Leverage to Paradigm was much more like CEA vs GWWC (two brands of the same organization) or even CEA UK vs CEA USA (two organizations acting together as one brand) than CEA vs ACE (one organization that spun off another one, which is now operates entirely independently with no overlap of staff etc).


Comment author: Geoff_Anders 05 August 2018 03:52:50AM 13 points [-]

Hi Jeff,

Sure, happy to try to clarify. I run both Leverage and Paradigm. Leverage is a non-profit and focuses on research. Paradigm is a for-profit and focuses on training and project incubation. The people in both organizations closely coordinate. My current expectation is that I will eventually hand Leverage off while working to keep the people on both projects working together.

I think this means we’re similar to MIRI/CFAR. They started with a single organization which led to the creation of a new organization. Over time, their organizations came to be under distinct leadership, while still closely coordinating.

To understand Leverage and Paradigm, it’s also important to note that we are much more decentralized than most organizations. We grant members of our teams substantial autonomy in both determining their day-to-day work and with regard to starting new projects.

On residence, new hires typically live at our main building for a few months to give them a place to land and then move out. Currently less than 1/3 of the total staff live on-site.

Comment author: Jeff_Kaufman 06 August 2018 12:27:08PM 21 points [-]

Thanks for clarifying!

Two takeaways for me:

  • Use of both the "Paradigm" and "Leverage" names isn't a reputational dodge, contra throwaway in the original post. The two groups focus on different work and are in the process of fully dividing.

  • People using what they know about Leverage to inform their views of Paradigm is reasonable given their level of overlap in staff and culture, contra Evan here and here.

Comment author: weeatquince  (EA Profile) 05 August 2018 09:57:53PM *  25 points [-]

counting our research as 0 value, and using the movement building impact estimates from LEAN, we come out well on EV compared to an average charity ... I will let readers make their own calculations

Hi Geoff. I gave this a little thought and I am not sure it works. In fact it looks quite plausible that someone's EV (expected value) calculation on Leverage might actually come out as negative (ie. Leverage would be causing harm to the world).

This is because:

  • Most EA orgs calculate their counterfactual expected value by taking into account what the people in that organisation would be doing otherwise if they were not in that organisation and then deduct this from their impact. (I believe at least 80K, Charity Science and EA London do this)

  • Given Leverage's tendency to hire ambitious altruistic people and to look for people at EA events it is plausible that a significant proportion of Leverage staff might well have ended up at other EA organisations.

  • There is a talent gap at other EA organisations (see 80K on this)

  • Leverage does spend some time on movement building but I estimate that this is a tiny proportion of the time, >5%, best guess 3%, (based on having talked to people at leverage and also based on looking at your achievements to date compared it to the apparent 100 person-years figure)

  • Therefore if the amount of staff who could be expected to have found jobs in other EA organisations is thought to be above 3% (which seems reasonable) then Leverage is actually displacing EAs from productive action so the total EV of Leverage is negative

Of course this is all assuming the value of your research is 0. This is the assumption you set out in your post. Obviously in practice I don’t think the value of your research is 0 and as such I think it is possible that the total EV of Leverage is positive*. I think more transparency would help here. Given that almost no research is available I do think it would be reasonable for someone who is not at Leverage to give your research an EV of close to 0 and therefore conclude that Leverage is causing harm.

I hope this helps and maybe explains why Leverage gets a bad rep. I am excited to see a more transparency and a new approach to public engagement. Keep on fighting for a better world!

*sentence edited to better match views

Comment author: Paul_Crowley 04 August 2018 10:35:43PM 15 points [-]

Could you comment specifically on the Wayback Machine exclusion? Thanks!

Comment author: Khorton 03 August 2018 11:15:21PM 5 points [-]

What have you done to promote movement building? I didn't see anything on the post or your website, other than the summit next week.

Comment author: Geoff_Anders 04 August 2018 06:12:46AM *  19 points [-]


(1) founded THINK, the first EA student group network

(2) ran the EA Summit 2013, the first large EA conference (video)

(3) ran the EA Summit 2014

(4) ran the EA Retreat 2014, the first weeklong retreat for EA leaders

(5) handed off the EA Summit series to CEA; CEA renamed it EA Global

(6) helped out operationally with EA Global 2015.

Comment author: Dunja 04 August 2018 02:23:52PM 3 points [-]

Part of what we do is help people to understand themselves better via introspection and psychological frameworks.

Could you please specify which methods of introspection and psychological frameworks you employ to this end, and which evidence you use to assure these frameworks are based on the adequate scientific evidence, obtained by reliable methods?