markov_user

Karma: 410

markov_user Jul 10, 2022, 2:28 PM
2 points
0 ∶ 0
in reply to: Soren’s comment on: One’s Future Behavior as a Domain of Calibration
Sort of, so firstly I have a field next to each prediction that automatically computes its “bucket number” (which is just FLOOR(<prediction> * 10)). To then get the average probability of a certain bucket, I run the following: =AVERAGE(INDEX(FILTER(C$19:K, K$19:K=A14), , 1)) - note that this is google sheets though and I’m not sure to which degree this transfers to Excel. For context, column C contains my predicted probabilities, column K contains the computed bucket numbers, and A14 here is the bucket for which I’m computing this. Similarly I count the number of predictions of a given bucket with =ROWS(FILTER(K$19:K, K$19:K<>"", K$19:K=A14)) and the ratio of predictions in that bucket that ended up true with =COUNTIF(FILTER(D$19:K, K$19:K=A14), "=1") / D14 (D19 onwards contains 1 and 0 values depending on if the prediction happened or not; D14 is the aforementioned number of predictions in that bucket).
If this doesn’t help, let me know and I can clear up one such spreadsheet, see if I can export it as xlsx file and send it to you.

markov_user Feb 12, 2022, 9:31 PM
1 point
0 ∶ 0
on: Reimagining a social network for EA community
Thanks for sharing! I’ve had the feeling for a while that it would be great if EA managed to make goals/projects/activities of people (/organizations) more transparent to each other. E.g. when I’m working on some EA project, it would be great if other EAs who might be interested in that topic would know about it. Yet there are no good ways that I’m aware of to even share such information. So I certainly like the direction you’re taking here.
I guess one risk would be that, however easy to use the system is, it is still overhead for people to have their projects and goals reflected there. Unless it happens to be their primary/only project management system (which however would be very hard to achieve).
Another risk could be that people use it at first, but don’t stick to it very long, leading to a lot of stale information in the system, making it hard to rely on even for highly engaged people.
I guess you could ask two related questions. Firstly, let’s call it “easy mode”: assuming the network existed as imagined, and most people in EA were in fact using this system as intended—would an additional person that first learns of it start using it in the same productive way?
And secondly, in a more realistic situation where very few people are actively using it, would it then make sense for any single additional person to start using it, share their goals and projects, keep things up to date persistently, probably with quite a bit of overhead on their part because it would happen on top of their actual project management system?
I think it’s great to come up with ideas about e.g. “the best possible version EA Hub” and just see what comes out, even though it’s hard to come up with ideas that would answer both above questions positively. Which is why improving the EA Hub generally seems more promising to me than building any new type of network, as at least you’d be starting with a decent user base and would take away the hurdles of “signing up somewhere” and “being part of multiple EA related social networks”.
So long story short, I quite like your approach and the depth of your mock-up/prototype, and think it could work as inspiration for EA Hub to a degree. Have my doubts that it would be worthwhile actually building something new just to try the concept. Except maybe creating a rough interactive prototype (e.g. paper prototype or “click dummy”), and playing it through with a few EAs, which might be worthwhile to learn more about it.

markov_user Feb 10, 2022, 8:45 PM
3 points
0 ∶ 0
on: Low-Commitment Less Wrong Book (EG Article) Club
I’d be up for the reading and comment writing part (will see if it works out time-wise), probably not so much for zoom. Nice idea and thanks for taking the initiative!

markov_user Feb 7, 2022, 8:45 PM
11 points
0 ∶ 0
on: Stop procrastinating on career planning
Is your post deliberately categorized as question? The four questions included in it all seem to be of the rhetorical kind. :P
Thanks for the post though! I think I’m in a very similar situation and you basically convinced me. I didn’t expect five minutes ago to be just one minute of reading away from being convinced of applying to a 80000 hours career advice call, yet here we are.

markov_user Jan 29, 2022, 9:46 PM
5 points
0 ∶ 0
on: Giving Multiplier after 14 months
Great write-up! The “many people are happy to donate to effective charities as long as they also donate to their favorite charity” point did indeed come as a surprise. Seems like a very valuable insight for certain types of outreach.

markov_user Jan 9, 2022, 10:12 AM
3 points
0 ∶ 0
in reply to: PeterSlattery’s comment on: Introducing Effective Self-Help
I think that you should consider connecting and collaborating with key parties who have interdependent goals & similar incentives
A small addition to your list would be this post about a study on a depression related intervention that I believe originated from within the EA community. Might well be worth contacting the author.

markov_user Jan 9, 2022, 10:06 AM
2 points
0 ∶ 0
on: Introducing Effective Self-Help
Interesting project! It reminds me a bit of Huberman lab, the existence and apparent popularity of which could be taken as an argument in favor of ESH to be worthwhile (although format, target audience and focus might of course differ quite a bit).
One thing I personally find very interesting is the point you mentioned as a counter argument: “Individual differences in benefit significantly outweigh the general differences in value between interventions”—in my opinion, this could even be viewed as quite the opposite: my impression is that in most easily digestible sources (such as pop-sci books, podcasts, blogs), this point is mostly ignored, and getting reliable information about this facet of health and well-being interventions would be great.
People very often speak of effect sizes as if they were a thing inherent to an intervention or substance, when actually they quite often seem to depend strongly on the person. An intervention with a strong positive effect size on only 10% of people could be much more exciting than an intervention with a weak effect on everybody. Even a large positive effect on small number of people combined with negative effects on the rest could be a very useful intervention, given you find out early enough whether it works for you or not. Getting some insight into the nature of variance of different interventions, if such data is available, could be really useful. It might of course be the case that most studies don’t offer such insights, because it’s impossible to tell whether the subset of participants that benefited from an intervention can be attributed to noise or not.

markov_user Dec 25, 2021, 6:47 AM
2 points
0 ∶ 0
on: Longtermism in 1888: fermi estimate of heaven’s size.
This is great, thanks for sharing!
I found the “let’s assume humanity remains at a constant population of 900 million” notion particularly interesting. On some level I still have this (obviously wrong) intuition that human knowledge about its history just grows continuously based on what happens at any given time. E.g. I would have implicitly assumed that a person living in 1888 must have known how the population numbers have developed over the preceding centuries. This is of course not necessarily the case for a whole bunch of reasons, but seeing that he wasn’t even aware that population growth is a thing was a serious surprise (unless he was aware, but thought it was close enough to the maximum to be negligible in the long term?).
It’s funny how he assumes a generation would span 31.125 years without giving any explanation for that really specific number. Maybe he had 8 children at this point in time, and took e.g. his average age during the birth of all of them?
And lastly, he as well as any readers of this letter would have greatly benefited of the scientific notation. Which makes me wonder what terrible inefficiencies in communication & encoding / expressing ideas we’re suffering from today, without having any inkling that things could be better… :)

markov_user Nov 15, 2021, 1:34 PM
4 points
0 ∶ 0
on: Preprint is out! 100,000 lumens to treat seasonal affective disorder
Nice! :)
Also, I think a few links are missing here:
David Chapman for inspiring us with these two posts in the Meaningness blog, Raemon for inspiring us with this LessWrong post

markov_user Nov 8, 2021, 4:11 PM
5 points
0 ∶ 0
on: Could EA be ideas constrained?
Some thoughts (not to say ideas) regarding 3:
- come up with more ideas
  - just brainstorming in a very unconstrained way on relevant questions (e.g. “babble”)
  - trying some systematic ways to identify implicit assumptions in our existing beliefs and ideas, and questioning them
  - looking at existing entities (orgs, fields, causes, tools...) and thinking about how they could be different
- share ideas more effectively in the movement
  - encourage sharing in the first place (makes me sad to read of posts people started in the past but never finished)
  - good compression of ideas (e.g. short posts, descriptive titles, beginning with a summary)
  - make things easy to find via search
  - talk to other EAs about your ideas
  - get feedback early on
    maybe twitter is good for this?
- actual implementation
  - a lot of ideas may exist, e.g. in the dusty archives of this forum, that nobody has ever acted on and people have more or less forgotten about or never heard of in the first place
  - some (or many) people may generally be more interested in thinking—maybe EA is implementation constrained rather than idea constrained after all? (but I guess there are a lot of constraints anyway, and they vary substantially by who you ask; so idea constraints most certainly are a thing, affecting some more than others)

markov_user Nov 8, 2021, 3:46 PM
4 points
0 ∶ 0
in reply to: PeterSlattery’s comment on: Could EA be ideas constrained?
There are definitely many coincidence of wants related problems, where someone has a good idea that someone would do or fund but that person never hears of it.
Very much agree with your points, this one in particular. I think in a perfect world we would all have a way of knowing of what others in the EA community are thinking about, working on and what they need help with. I’d love to have a way to share more openly (but without wasting other’s attention) what I’m focusing on so that others who think about similar things could be made aware of this opportunity for collaboration. But I don’t really know of any practical ways to achieve this. Write a forum post saying “Hey everyone I’m really interested in X recently and plan to spend the next 3 months diving into that topic”? Probably not.
EA G(X) could be helpful, because you can share your (current) interests in your profile on the networking app. And theoretically find others who mention the same keywords. But then swapcard comes along and doesn’t support proper searching, so I missed out on many potentially great relevant contacts. :( Plus of course it doesn’t happen all that often, and always contains only a relatively small subset of the community.

markov_user Nov 8, 2021, 3:28 PM
3 points
0 ∶ 0
in reply to: Aaron Gertler 🔸’s comment on: What high-level change would you make to EA strategy?
One thing I could imagine being very helpful is some kind of ongoing local group “mentoring”. So instead of one or two single calls on strategy or bottlenecks, having some experienced person more deeply invested with any particular local group in need. Somebody who might (occasionally) participate at our virtual meetups, our planning/strategy calls, gets to know our core members, our situation, needs and problems, and can provide actionable insights on all of them.
The problem with calls I’ve had in the past is that it’s quite difficult to get accross everything relevant, so we might just focus on one or two issues, obtain some pointers to other people or resources who might be helpful, and some relatively generic advice. Not to say it isn’t useful—but it also doesn’t seem to be like a complete solution. I’ve also read most of the EA Hub resources on running a group, but tend to come out of these articles thinking “yup this makes sense” and not actually turning it into anything concrete. Which, again, is probably entirely my responsibility. But I could imagine I’m not the only time/energy constrained local group coordinator struggling with properly utilizing the existing resources.
On the other hand, such more involved support over longer time of course also comes with significantly higher cost, and I can’t tell whether that would be worth it.

markov_user Nov 8, 2021, 2:50 PM
4 points
0 ∶ 0
on: Disagreeables and Assessors: Two Intellectual Archetypes
The distinction reminds me of the foxes vs hedgehogs model from Superforecasting / Tetlock. Hedgehogs being “great idea thinkers” seeing everything in the light of that one great idea they’re following, whereas foxes are more nuanced, taking in many viewpoints and trying to converge on the most accurate beliefs. I think he mentioned in the book that while foxes tend to make much better forecasters, hedgehogs are not only more entertaining but also good in coming up with good questions to forecast in the first place.
An entirely different thought: The Laws of Human Nature by Robert Greene was the first audible book I returned without finishing. It was packed with endless “human archetypes” described in great detail, making some rather bold claims about what “this type” will do in some given situation. You mention in the footnotes already that people who dislike e.g. personality profiling tools might not like this post. And it did indeed somewhat remind me of that book, but maybe your “assessor” way of describing the model, as opposed to Greene’s very overconfident seeming way of writing, made this seem much more reasonable. There seems to be a fine line between actually useful models of this kind which have some predictive power (or at least allow thoughts to be a bit tidier), and those that are merely peculiarly entertaining, like Myers-Briggs. And I find it hard to tell from the outside on which side of that line any given model falls.

markov_user Nov 8, 2021, 2:28 PM
3 points
0 ∶ 0
on: [Discussion] How do you convince skeptics of AI safety?
One thing I could imagine happening in these situations is that people close themselves off to object level arguments to a degree, and maybe for (somewhat) good reason.
- to the general public, the idea of AI being a serious (existential) risk is probably still very weird
- people may have an impression that believing in such things correlates with being gullible
- people may be hesitant towards “being convinced” of something they haven’t fully thought through themselves
I remember once when I was younger talking to a Christian fanatic of sorts, who kept coming up with new arguments for why the bible must obviously be true due to the many correct predictions it has apparently made, plus some argument about irreducible complexity. In the moment, I couldn’t really tell if/where/why his arguments failed. I found them somewhat hard to follow and just knew the conclusion would be something that is both weird and highly unlikely (for reasons other than his concrete arguments). So my impression then was “there surely is something wrong about his claims, but in this very moment I’m lacking the means to identify the weaknesses”.
I sometimes find myself in similar situations when some person tries to get me to sign something or to buy some product they’re offering. They tend to make very convincing arguments about why I should definitely do it. I often have no good arguments against that. Still, I tend to resist many of these situations because I haven’t yet heard or had a chance to find the best counter arguments.
When somebody who has thought a lot about AI safety and is very convinced of its importance talks to people to whom this whole area is new and strange, I can imagine similar defenses being present. If this is true, more/better/different arguments may not necessarily be helpful to begin with. Some things that could help:
- social proof (“these well respected people and organizations think this is important”)
- slightly weaker claims that people have an easier time agreeing with
- maybe some meta level argument about why the unintuitive-ness is misguided (although this could probably also taken as an attack)

markov_user Nov 8, 2021, 7:50 AM
1 point
0 ∶ 0
on: Best practices for organizational effectiveness
The cost of this seems pretty low, but in a way the expected value too seems limited (to me at least from the context you provided): I’d assume that unless this turns out to be so good that it becomes a “standard” of sorts (that people always tend to mention whenever organizational ineffectiveness comes up), it would likely end up as a relatively short lived project that doesn’t reach too many people and organizations. Although this could partially be mitigated if it’s stored in a persistent, easy to search and find way, so that future people on the lookout for such a guide would stumble upon it and immediately see its value.

markov_user Sep 20, 2021, 9:23 PM
2 points
0 ∶ 0
on: Introducing TEAMWORK—an EA Coworking Space in Berlin
Sounds really cool. Time to visit Berlin! :)

markov_user Sep 14, 2021, 7:44 PM
1 point
0 ∶ 0
on: Learning, Knowledge, Intelligence, Mastery, Anki—TYHTL post 2
Just a side note: While Obsidian is free (and great), I’m pretty sure it’s not open source.

markov_user Sep 11, 2021, 12:38 PM
1 point
0 ∶ 0
in reply to: mic’s comment on: Lessons from Running Stanford EA and SERI
Thank you Michael!
- I personally am definitely more time- than funding constrained. Or maybe evem “energy constrained”? But maybe applying for funding would be something to consider when/if we find a different person to run the local group, maybe a student who could do this for 10h a week or so.
- regarding a fellowship: my bottlenecks here are probably “lack of detailed picture of how to run such a thing (or what it even is exactly)” and “what would be the necessary concrete steps to get it off the ground”. Advertising is surely very relevant, but secondary to these other questions for now.
- on a slightly more meta level, I think one of the issues is that I don’t have a good overview of the “action space” (or “moves”) in front of me as an organizer of an EA local group. Running a fellowship appears to be a very promising move, but I don’t really know how to make it. Other actions may be intro talks, intro workshops, concepts workshops, discussions, watching EAG talks together, game nights, talks in general, creating a website, setting up a proper newsletter instead of having a manually maintained list of email addresses, looking for a more capable group organizer, facebook ads, flyers, posters, running giving games, icebreaker sessions, running a career club, coworking, 1on1s, meeting other local groups, reaching out to formerly-but-not-anymore-active members, and probably much more I’m not even thinking about. Maybe I’m suffering a bit from decision paralysis here and just doing any of these options would be better than my current state of “unproductive wondering what I should be doing”… :)
- will message you regarding a call, thanks for the offer!

markov_user Sep 4, 2021, 8:27 PM
2 points
0 ∶ 0
on: If you could send an email to every student at your university (to maximize impact), what would you include in it?
Given I just received a link to this article in the 80,000 Hours newsletter: https://80000hours.org/make-a-difference-with-your-career/ -- that article seems like something that a lot of students might potentially be interested in. So something like a brief description of the key idea plus a link to the article would be one option.

markov_user Sep 4, 2021, 11:39 AM
4 points
0 ∶ 0
on: A Case for Better Feeds
Recently I’ve been thinking a lot about the flow and distribution of information (as in facts/ideas/natural language) as a meta level problem. It seems to me that “ensuring the most valuable information finds its way to the people who need it” could make a huge difference to a lot of things, including productivity, well-being, and problem-solving of any kind, particularly for EAs. (if anybody reading this is knowledgeable in this broad area, please reach out!)
Your post appears to focus on a very related issue, which is how EAs source their EA information and some specific ways to improve it. I definitely agree that this is an issue worth looking into and worth improving (I personally think that either the EA forum or the EA Hub are in the best position to make such improvements, although I’m unsure what these improvements would look like).
The EA Forum Job Hunt idea admittedly doesn’t seem very promising to me from how I understood it—it sounds like by far the most work of all the suggestions, for a problem that, to me, seems as if it’s solved to a pretty reasonable degree.
I don’t quite understand the EA Hub suggestion. What would be submitted and upvoted? Just the existence of (local) groups?
The remaining points regarding Twitter bots and feeds sound good to me, simply because they sound like very little work (unless I’m misjudging that), while potentially being helpful to probably many dozens of EAs.
By the way I do wonder what ratio of EAs is actively using twitter. I for once am not at all, and am not aware of many people I know personally doing so, but that might not mean much and not be very representative.