If you agree it is a serious and baseless allegation, why do you keep engaging with him? The time to stop engaging with him was several years ago. You had sufficient evidence to do so at least two years ago, and I know that because I presented you with it, e.g. when he started casually throwing around rape allegations about celebrities on facebook and tagging me in the comments, and then calling me and others nazis. Why do you and your colleagues continue to extensively collaborate with him?

To reiterate, the arguments he makes are not sincere: he only makes them because he thinks the people in question have wronged him.

I think that X-risk reduction projects also have a problem with measurement!

However, measuring the extent to which you’ve reduced X-risk is a lot harder than measuring whether students have taken some kind of altruistic action: in the latter case, you can just ask the students (and maybe give them an incentive to reply).

Thus, if someone wants me to donate to their “EA education project”, I’m probably going to care more about direct outcome measurement than I would if I were asked to support an X-risk project, because I think good measurement is more achievable. (I’d hold the X-risk project to other standards, some of which wouldn’t apply to an education project.)

“Is EA Growing? EA Growth Metrics for 2018” has some data on this, and I look forward to doing it again for 2019-2020

It’s currently very unclear where to find the Wiki content. For example, I expected to find it here and also here. Reading this discussion (by Pablo and Michael), I expected to see the Wiki page being discussed, or at least a link to it, but I couldn’t.

The “all tags” page will eventually (within the next few weeks) link to all articles, including those that are currently “wiki only”.

The wiki-only version of the tag being discussed was merged with the active tag, so the wiki-only version no longer exists (see subsequent posts in the discussion).

I think the idea that you’re presenting Wiki pages as modified versions of “tags” is very (and fundamentally) confusing user experience (even if that’s how they have to be coded). Can users not understand wiki pages as first-class entities, and say that a tag is one way that we can use a Wiki page?

That’s how we’ve decided to refer to those entities internally (the wiki as a collection of articles, some of which are also tags). This decision was made over the weekend, so previous discussion of the wiki is a bit of a hodgepodge — we’ll use the less confusing system in future discussions (and try to enshrine it in the UI as well).

80K has a deep network of contacts who send them jobs, and has a staff member who spends many hours/​week keeping their job board up to date. I think it would be very difficult to find enough volunteers to replicate that effort on the wiki (especially for things that change as quickly as job openings). If I could find that many dedicated volunteers, I’d prefer to have them work on things that an org with 80K’s resources isn’t already doing.

This is a good suggestion — I’ll pass it on to the developers.

All tags are already “editable all the time”, other than a few we’ve locked as “admin only” for various reasons (e.g. the “frontpage” tag, which isn’t really an article). The editing process is the same as for Forum posts — very easy!

If you’re referring to posts rather than tags, that’s a very different suggestion.

I could imagine this being a future feature! That facet of Wikipedia is well worth imitating.

Note on this response: I really appreciate your engagement on this! My goal for this comment is to clarify some things I didn’t go into much detail on before, and better represent the way we’re currently thinking about the project (as something CEA cares about a lot and will continue to care about).

That said, I’m still pretty sure there’s a lot of low-hanging fruit in terms of content that’s notable, EA relevant, and not on Wikipedia. “Longtermism” is a good example. I also suspect (though more experienced Wikipedia people should weigh in) that if we made good progress on the low hanging fruit, terms like “hinge of history” and “patient philanthropy” would be perceived as considerably more notable than they are now.

I agree with all of this. However, I think CEA would have to tread carefully to support this work without violating Wikipedia’s rules about paid editing. I may think about this more in future months (right now, I’m juggling a lot of projects). If you have suggestions for what CEA could do in this area, I’d be happy to hear them.

Meanwhile, I’ll declare this as clearly as I can, using bold text in lieu of cash: I am very happy to see work done on EA-relevant Wikipedia pages, and I think that such work ought to be appreciated by the community at large (and, where appropriate, considered in grant applications, job applications, etc.)

While I don’t have experience as a Wikipedia editor, I have quite a bit of experience with a previous employer’s internal Wiki. It was immensely valuable. It was also immensely difficult to develop and maintain. There were countless emails and meetings on the theme “we need to clean up the Wiki (and this time we really mean it!)”, and in my experience it takes a combination of that and making the Wiki part of peoples job responsibilities/​evaluations to maintain something useful. It’s amazing how quickly information gets stale, and it gets harder to keep things updated the as you get more entries in your Wiki.

I also have experience with an employer’s (valuable) internal wiki, and I appreciate this point. However, I’d expect that keeping information extremely up-to-date (e.g. making weekly updates to a large range of entries) is going to be more important for a corporate wiki than a conceptual wiki.

My employer’s internal wiki had lots of articles that were constantly becoming wrong, in ways that would impede our work if the wrongness wasn’t corrected (“this is no longer the password you need”, “this menu item has been renamed”, etc.). On the other hand, articles like “Longtermism” or “Wild Animal Suffering”, may be expanded from time to time, but it’s rare that text in such an article will suddenly become wrong.

This doesn’t mean that decay isn’t a concern — just that it’s less of a crisis than it would be if e.g. a company were to stop making it anyone’s responsibility to edit their own wiki.

If you build an EA Wiki that gets to the level of having a page for “Russia” (to use Aaron’s example), you’re going to need a lot more volunteer (and/​or paid staff) time than a few days of someone’s time a month.

I think I wasn’t clear enough in what I meant the “15-20 hours” to represent, and I may have come off as blasé in a way I didn’t intend.

Quoting myself:

I’d expect the bare minimum to be less time than Pablo and I are putting in, such that further volunteer contributions are a nice addition rather than an existential necessity.

To clarify: I think of “bare minimum” as something like “many articles are checked on ~once per year, and the volunteer spends a few minutes thinking about whether they want to add anything; a volunteer looks at each edit made by a Forum user and makes small fixes/​reversions as needed”.

With “many”, I’m leaving out articles like “decision theory” and “Abhijit Banerjee”, and focusing on e.g. articles about core EA orgs and active research areas.

And when I think of “volunteer time”, I’m thinking of people who see themselves as “wiki volunteers”, rather than the general population of the Forum — I’d also expect people on the Forum to put in quite a bit more time, because (unlike those of other EA wikis) the Forum’s wiki articles will be very salient to them.

Our tag pages got over 2000 views in each of January and February, even before most of the content uploading happened. By the time we’ve put in months of additional paid work and added new features to improve the wiki and make it more visible, I expect that number to increase by a lot. Only a tiny fraction of those views will turn into edits, whatever UI we employ to encourage them, but that’s still a lot of additional hours.

Still, I wouldn’t think of a user who makes one five-minute edit every six months as a “volunteer”. So my 15-20 hour estimate didn’t include this kind of activity.

The result I’d expect from a “bare minimum” outcome, combining dedicated volunteer work and other user edits: the wiki continues to be a useful resource which people refer to often, especially for its articles on evergreen concepts that don’t change often. It does decay somewhat, and few new articles are written, but it remains substantial enough that a few dedicated people could jump in and resurrect it more easily than they could start a new wiki.

It’s possible that 15-20 hours/​month is too low even to expect this level of maintenance; as I note below, I didn’t spend much time coming up with the estimate (as I don’t think it will be relevant for a long time).

*****

That said, I would be very disappointed in the “bare minimum” outcome, and would go to considerable lengths to bring in more volunteer support. I mostly set my own priorities at CEA; even if I came to believe that doing a lot of dedicated wiki work wasn’t a good use of my time, and we decided to stop paying for work from Pablo or others like him, I can’t imagine not wanting to spend some of my time coordinating other people to do this work.

A number I’d be happy with, where I’d expect to see the wiki grow and flourish? At 500 articles, with three hours/​article/​year, that’s 125 hours/​month — not a very confident estimate, but one that seems plausible for a good outcome. That would take a lot of coordination (if we want to have a couple dozen people putting in an hour per week*), but I’d expect to be on the front lines of that effort.

*I’d expect the actual structure of editing to look a bit different than this, because so many members of the community are going to be invested in specific articles; I’d guess that e.g. someone from 80K would make substantial edits to 80K’s article every so often, and that the same would be true for many other editor/​topic combinations.

My overarching concern is that you’re seriously underestimating the ongoing costs of this project, which will basically continue in perpetuity and increase over time. This has been the issue that sank previous attempts at an EA Wiki, and honestly it’s a pretty big red flag that you “don’t have a great estimate for how much volunteer time we’d need to keep things running.”

The reason I haven’t spent much time thinking about the “volunteer-only” version of the wiki is that Pablo has a grant to work on this project for many months to come, and the project is also one of my highest current priorities at CEA. If it starts to seem like one or both of those things will stop being true in the foreseeable future, I expect to put a lot more time into preparing for the “volunteer-only” era.

A comparison: If you asked me “Aaron, who would take over the EA Newsletter if you got hit by a bus?”, I wouldn’t have a good answer on hand. That doesn’t mean I think the EA Newsletter isn’t important, or that it doesn’t take much time to produce; I just don’t expect to stop running it anytime soon, or to be hit by a bus.

Wikipedia already has more features (including features that would be valuable for an EA Wiki like translation) and a much bigger team (that doesn’t have competing priorities and doesn’t cost CEA anything), so it seems to me like CEA/​LW will be constantly playing catchup. And any time they spend adding new features or even just maintaining a dedicated Wiki is time they won’t be able to work on other valuable projects.

As I’ve said, I really want to see people contribute to the real Wikipedia in addition to our dedicated wiki.

But given that LessWrong already has their own wiki, which ours is a copy of, I expect them to keep adding new features to theirs which ours will adopt (this is already how the EA Forum gets most of its new features). This is how they’d be spending their time with or without us.

I’ll retract my “toss-up” comment; I don’t really know what I meant by it, having written that comment quickly and without editing. I do think Wikipedia will always have a better overall feature set than our Wiki — thanks for making that point clearly.

But I expect us to occasionally implement things that are a bit better than Wikipedia’s version of the thing, and I also think there’s a lot of value in having something that is almost as good as Wikipedia in most basic respects that we also control (rather than being at the mercy of notoriously hard-headed admins).

Just wanted to quickly hop in to say that I think this little sub-thread contains interesting points on both sides, and that people who stumble upon it later may also be interested in Forum posts tagged “epistemic humility”.

Cool idea! I’m not doing a PhD, so I won’t partake. But I previously wrote a comment intended to summarise Wiblin and Greenberg’s discussion of this sort of idea, and my own thoughts on it, which I’ll copy below in case that’s useful for anyone reading this post. (Obviously this overlaps somewhat with what you already say in your post.)

---

People would often benefit from more “line management”, and could often get it just by setting up weekly meetings with someone else who’s in a similar boat

A chunk of the interview is devoted to these points. From memory, some points made were:

• It can be weirdly useful to have ~weekly meetings to just discuss what one did last week, what went well and poorly, what one’s goals are, and what one plans to do next week

  • The reason the usefulness is weird/​surprising is that a lot of the benefit seems to come from just having these meetings at all and being asked obvious questions like “Would that task really be the best thing to do to achieve your goals?”

    • And in theory, one could just simulate these conversations by oneself

    • But at least for many people, it seems to be more effective to have an actual conversation with another person

  • This can help with both productivity (including getting the right tasks done) and mood (e.g., reducing self-doubt or a sense of listlessness)

• Some people don’t have someone to provide that “line management” role

  • E.g., people in PhD programs might not get frequent enough meetings with their advisor, or it might be clear that their advisor doesn’t care about or is terrible at management

• Those people might benefit from just arranging to have weekly meetings with a friend, colleague, fellow student, or similar who’s in a similar boat

  • Having weekly meetings with the same person allows them to have more context on one’s full situation, goals, skills, etc., which seems helpful for this

  • PhD students could arrange this themselves, and it might help combat PhD programs often seeming to be unusually crushing experiences (due to a lack of guidance, feedback, etc.)

  • (Something I’m not sure they explicitly said, but which seems true: This could probably be useful even for people who do have meetings with a line manager, if the person would benefit from meeting more, or if the manager sucks.)

That first point resonated with me very much. I received basically no line management at the school I taught at, and that sucked, and I only realised how much it sucked once I moved into roles where I did have weekly line management style meetings and discovered how helpful they were (both for my productivity and for my mood).

And that third point seemed like an obvious but great idea. I intend to apply it myself if I find myself in a future situation where it’s relevant. And I intend to keep it in mind as something to maybe suggest to people, when it seems relevant.

[This comment is a tangential and clarifying question; I haven’t yet read your post]

Ord’s book is a great restatement of the ethical case, though I disagree with his prioritisation of climate change, nuclear weapons and collapse

If I didn’t know anything about you, I’d assume this meant “Toby Ord suggests climate change, nuclear weapons, and collapse should be fairly high priorities. I disagree (while largely agreeing with Ord’s other priorities).”

But I’m guessing you might actually mean “Toby Ord suggests climate change, nuclear weapons, and collapse should be much lower priorities than things like AI and biorisk (though they should still get substantial resources, and be much higher priorities than things like bednet distribution). I disagree; I think those things should be similarly high priorities to things like AI and biorisk.”

Is that guess correct?

I’m not sure whether my guess is based on things I’ve read from you, vs just a general impression about what views seem common at CSER, so I could definitely be wrong.

I highly recommend the Bodhicaryavatara by Shantideva! It’s the most significant ethical text of the Mahayana Buddhist tradition, with some serious Madhyamaka metaphysics sprinkled in. I’m currently writing my undergrad thesis on it, and I’d be happy to talk about it.

Here’s a great guide: https://​​www.shambhala.com/​​guide-to-the-way-of-the-bodhisattva/​​. I took an intensive course on the Bodhicaryavatara in the traditional monastic style in Kathmandu, Nepal; see https://​​ryi.org/​​programs/​​degree-programs if you really want to dive deep. The school is currently offering all courses online.

I’m also studying The Ladder of Divine Ascent by St. John Climacus, with the accessible commentary Thirty Steps to Heaven by Vassilios Papavassiliou. St. John is venerated in both the Eastern Orthodox Church and the Roman Catholic Church, but he gets more attention from the Orthodox. I find Orthodoxy fascinating because it has such a mystical relationship-oriented spirituality compared to the legalistic style of both Catholicism and Protestantism. However, this text focuses on individual spirituality; there’s not much discussion of ethics.

Thanks, Haydn, for writing this thoughtful post. I am glad that you (hopefully) found something from the syllabus useful and that you took the time to read and write about this essay.

I would love to write a longer post about Torres’ essay and engage in a fuller discussion of your points right away, but I’m afraid I wouldn’t get around to that for a while. So, as an unsatisfactory substitute, I will instead just highlight three parts of your post that I particularly agreed with, as well as two parts that I believe deserve further clarification or context.

A)

Torres suggests that longtermism is based on an ethical assumption of total utilitarianism (...) However, although total utilitarianism strongly supports longtermism, longtermism doesn’t need to be based on total utilitarianism.

I agree with this and think that any critique of longtermism’s moral foundations should engage seriously with the fact many of its key proponents have written extensively about moral uncertainty and pluralism, and that this informs longtermist thinking considerably. I don’t think Torres’ essay does that.

B)

However, the more common longtermist policy proposal is differential technological development – to try to foster and speed up the development of risk-reducing (or more generally socially beneficial) technologies and to slow down the development of risk-increasing (or socially harmful) technologies.

Agreed, this seems like another important omission from the essay and one that is quite conspicuous given Bostrom’s prominent essay on the topic.

C)

Torres underplays the crucial changes Ord makes with his definition of existential risk as the “destruction of humanity’s potential” and the institution of the “Long Reflection” to decide what we should do with this potential. Long Reflection proponents specifically propose not engaging in transhumanist enhancement or substantial space settlement before the Long Reflection is completed.

As above, this seems like a critical omission

D)

Torres implies that longtermism is committed to a view of the form that reducing risk from 0.001% to 0.0001% is morally equivalent to saving e.g. thousands of present day lives. (...)

However, longtermism does not have to be stated in such a way. The probabilities are unfortunately likely higher – for example Ord gives a ¹⁄₆ (~16%) probability of existential risk this century – and the reductions in risk are likely higher too. That is, with the right policies (e.g. robust arms control regimes) we could potentially reduce existential risk by 1-10%.

Unless I’m misunderstanding something, this section seems to conflate three distinct quantities:

1. The estimated marginal effect on existential risk of some action EAs could take.

2. The estimated absolute existential risk this century.

3. The estimated marginal effect on existential risk of some big policy change, e.g. arms control.

While (2) might indeed be as high as ~16%, and (3) may be as high as 1-10%, both of these quantities are very different from (1). Very rarely, if ever, do EAs have the option ‘spend $50M to achieve a robust arms control regime’; it’s much more likely to be ‘spend $50M to increase the likelihood of such a regime by 1-5%.’

So, unless you think the tens of millions of “EA dollars” allocated towards longtermist causes reduce existential risk by >>0.001% per, say, ten million dollars spent, then it seems like you would indeed have to be committed to Torres’ formulation of the tiny-risk-reduction vs. current-lives-saved tradeoff.

Of course, you may believe that the marginal effects of many EA actions are, in fact, >>>0.001% risk reduction. And even if you don’t, the tradeoff may still be a reasonable ethical position to take.

I just think it’s important to recognise that that tradeoff does seem to be a part of the deal for x-risk-focused longtermism.

E)

Torres suggests that longtermism is committed to donating to the rich rather than to those in extreme poverty (or indeed animals). He further argues that this reinforces “racial subordination and maintain[s] a normalized White privilege.”

However, longtermism is not committed to donating (much less transferring wealth from poor countries) to present rich people.

For a discussion of this point, I think it is only fair to also include the quote from Nick Beckstead’s dissertation that Torres discusses in the relevant section. I include it in full below, for context:

“Saving lives in poor countries may have significantly smaller ripple effects than saving and improving lives in rich countries. Why? Richer countries have substantially more innovation, and their workers are much more economically productive. By ordinary standards—at least by ordinary enlightened humanitarian standards—saving and improving lives in rich countries is about equally as important as saving and improving lives in poor countries, provided lives are improved by roughly comparable amounts. But it now seems more plausible to me that saving a life in a rich country is substantially more important than saving a life in a poor country, other things being equal.” (Beckstead, 2013, quoted in Torres, 2021)

Here, I should perhaps note that while I’ve read parts of Beckstead’s work, I don’t think I’ve read that particular section, and I would appreciate hearing if there is a crucial piece of context that’s missing. Either way, I think this quote deserves a fuller discussion – I will, for now, simply note that I certainly think the quote is very objectionable and potentially warrants indignation.

Again, thanks for writing the post, I look very much forward to the discussions in the comments!

My general intuition is that if there’s a strong case that some action today is going to make a huge difference for humanity dozens or hundreds of generations into the future, that case is still going to be pretty strong if we limit our horizon to the next 100 years or so.

I might be misunderstanding you here, so apologies if the rest of this comment is talking past you. But I think the really key point for me is simply that, the “larger” and “better” the future would be if we get things right,[1] the more important it is to get things right. (This also requires a few moral assumptions, e.g. that wellbeing matters equally whenever it happens.)

To take it to the extreme, if we knew with certainty that extinction was absolutely guaranteed in 100 years, then that massively reduces the value of reducing extinction risk before that time. On the other extreme, if we knew with certainty that if we reduce AI risk in the next 100 years, the future will last 1 trillion years, contain 1 trillion sentient creatures per year, and they will all be very happy, free, aesthetically stimulated, having interesting experiences, etc., then that makes reducing AI risk extremely important.

A similar point can also apply with negative futures. If there’s a non-trivial chance that some risk would result in a net negative future, then knowing how long that will last, how many beings would be in it, and how negative it is for those beings is relevant to how bad that outcome would be.

Most of the benefits of avoiding extinction or other negative lock-ins accrue more than 100 years from now, whereas (I’d argue) most of the predictable benefits of things like bednet distribution accrue within the next 100 years. So the relative priority of the two broad intervention categories could depend on how “large” and “good” the future would be if we avoid negative lock-ins. And that depends on having at least some guesses about the world more than 100 years from now (though they could be low-confidence and big-picture, rather than anything very confident or precise).[1]

So I guess I’m wondering whether you’re uncomfortable with, or inclined to dismiss, even those sorts of low-confidence, big-picture guesses, or just the more confident and precise guesses?

(Btw, I think the paper The Case for Strong Longtermism is very good, and it makes the sort of argument I’m making much more rigorously than I’m making it here, so that could be worth checking out.)

[1] If we’re total utilitarians, we could perhaps interpret “larger” and “better” as a matter of how long civilization or whatever lasts, how many beings there are per unit of time during that period, and how high their average wellbeing is. But I think the same basic point stands given other precise views and operationalisations.

[2] Put another way, I think I do expect that most things that are top priorities for their impact >100 years from now will also be much better in terms of their impact in the next 100 years than random selfish uses of resources would be. (And this will tend to be because the risks might occur in the next 100 years, or because things that help us deal with the risks also help us deal with other things.) But I don’t necessarily expect them to be better than things like bednet distribution, which have been selected specifically for their high near-term impact.

Another would be to add up the budgets of all EA and EA-recommended organizations (or EA-recommended programs).

I strongly second this

I second most of what Alex says here. Like him, I only know about this particular essay from Torres, so I will limit my comments to that.

Notwithstanding my own objections to its tone and arguments, this essay did provoke important thoughts for me – as well as for other committed longtermists with whom I shared it – and that was why I ultimately ended up including it on the syllabus. The fact that, within 48 hours, someone put in enough effort to write a detailed forum post about the substance of the essay suggests that it can, in fact, provoke the kinds of discussions about important subjects that I was hoping to see.

Indeed, it is exactly because I think the presentation in this essay leaves something to be desired that I would love to see more community discussion on some of these critiques of longtermism, so that their strongest possible versions can be evaluated. I realise I haven’t actually specified which among the essay’s many arguments that I find interesting, so I hope I will find time to do that at some point, whether in this thread or a separate post.

Would you include even cases that rely on things like believing there’s a non-trivial chance of at least ~10 billion humans per generation for some specified number of generations, with a similar or greater average wellbeing than the current average wellbeing? Or cases that rely on a bunch of more specific features of the future, like what kind of political systems, technologies, and economic systems they’ll have?

My general intuition is that if there’s a strong case that some action today is going to make a huge difference for humanity dozens or hundreds of generations into the future, that case is still going to be pretty strong if we limit our horizon to the next 100 years or so. Aside from technologies to prevent an asteroid from hitting the earth and similarly super-rare cataclysmic natural events, I’m hard pressed to think of examples of things that are obviously worth working on that don’t meet that test. But I’m happy to be further educated on this subject.

How do you feel about longtermist work that specifically aims at one of the following?

Yeah, that sort of “anti-fragile” approach to longtermism strikes me as completely reasonable, and obviously it has clear connections to the IIDM cause area as well.

A part of it, definitely. At the same time, there are other projects that may not offer much opportunity for innovation but where I still feel I can make a difference because I happen to be good at the thing they want me to do. So a more complete answer to your original question is that I choose and seek out projects based on a matrix of factors including the scale/​scope of impact, how likely I am to get the gig, how much of an advantage I think working with me would offer them over whatever the replacement or alternative would be, how much it would pay, the level of intrinsic interest I have in the work, how much I would learn from doing it, and how well it positions me for future opportunities I care about.