It's bad for all discussion topics to become inactive and ignored after around a day, like Reddit. The forum model, where topics go back to the top when anyone replies in them, is better for having longer, more serious or thoughtful discussions. Attention should be allocated to topics more based on writing activity instead of based on upvotes with a 1-2 day limit even for upvoted topics. Standard forums aren't a perfect system (e.g. they reward controversy) but they work better. I'm sure there's room for refinement of the forum model to help with problems if effort was put into a better system. But the basic current concept, that everything old gets ignored, is bad.
There's a big incentive to never write anything important or serious in comments, unless maybe a topic is new and popular (and it's hard to write important, serious things fast enough for that to work).
Interesting! I recently decided not to comment on a post that is not even a month old because there was no discussion about it anymore (and hadn't been even just a few days after it was posted) and I thought that my comment would be read by barely anyone. Not saying that this is generally a reasonable decision even in the current system, but sometimes it is even when one has something interesting to say... Maybe it's worth writing a longer post about this?
I don't know effective ways to cause EA to change, or, in the alternative, to get people to debate for the other side regarding my criticisms. And I don't want to play the social game of figuring out how to get the positive initial reactions, popularity and upvotes to draw more attention to my posts. I want an alternative truth-seeking game rather than to win more at the social game. But I don't know how to effectively say that to EA either.
Why not just use a standard forum like discourse or phpbb? (I think the answer is that a more social media style website was created on purpose. There are things people prefer about it, though they mostly don't like to state them and may not consciously know them.) Also curation would be more about social status than truth. Having some central gatekeepers deciding what ideas are seen by many people is really problematic (having it decided with voting is also really problematic, and btw I think EA's weighted voting is worse than one vote per person).
There are both positives and negatives to being addicted to the EA Forum, so I'm not sure I'm against it entirely. Also, I think the current gatekeeper(s) (mainly Lizka atm, I think) have better judgment than the aggregated wisdom of accumulated karma, so I really appreciate the curation. Weighted upvotes are controversial, and I could change my mind, but for now I suspect the benefits outweigh the costs because I trust the judgment of high-karma-power users more on average. For an alternative way to filter posts, I think Zoe William's weekly summaries are excellent.
Ideally, it would be cool if EA karma could be replaced with a distributed trust network of sorts, such as one I alluded to on LW.
Truth seeking requires things like arguments, criticism and debate – and people who take responsibility for doing truth seeking. Without attempts to organize those things and make them happen following standardized methods, I don't think EA will be very effective. Substitutes like curation or voting are never going to get similar results.
I don't know of standardised methods that I think are likely to more reliably generate that, and I am worried about rushing standardisation prematurely. But this worry is far from insurmountable, and I'd be happy to hear suggestions for things that should be standardised. You have any? : )
Yes I've been posting suggestions here and no one has engaged seriously or offered any substantive criticism, e.g. I posted about my debate policy, my suggestion of an anti-misquoting norm, my explanation that rationality policies parallel the rule of law, a long list of sample policies, and some explanation of how people should take individual responsibility for their ideas and it's very hard to get anywhere with a group like EA when it's no one's job or responsibility to deal with criticism so everyone can individually just ignore it. There's also a continuity problem where basically each individual tends to see a random selection of what I post, which makes it hard to explain anything long, complicated and based on some different background knowledge than most people here believe or are familiar with. I agree that standardizing methods is a work in progress rather than finished, but it seems most people are not working on it, don't want to, and don't want to debate the matter. Do you want to try to have a serious discussion over time and try to reach some conclusions?
Hmm, not sure how that job would work, but if someone could be paid to fix all of EAs coordination problems, that'd probably be worth the money. It is the responsibility of anyone who wishes to assume that responsibility. And if they can, I really hope they do.
I'm a very slow reader, but do you wish to discuss (or debate it!) over a video call sometime? I'm eager to learn things from someone who's likely to have a different background on the questions I'm interested in. : )
I doubt that's going to work, but I can ask some questions about your proposal to check. My first question is whether you had in mind a public or private call.
Mh, I had in mind both, and wanted to leave it up for interpretation. A public debate about something could be cool because I've never done that, but we'd need to know what we're supposed to disagree about first. Though, I primarily just wish to learn from you, since you have a different perspective, so a private call would be my second offer.
I consider calls inconvenient and inferior to writing, but I'll view it as a cultural difference and try to be flexible. So I'll do a recorded call, which I might put on YouTube, if we can agree on a time. Your listed times don't work for me but you seem to use this forum at later times. Would noon (US Pacific time) or later work for you? Nov 21?
I certainly take you seriously, Elliot. However, I am strapped for time so I have not been able to review your video or more of your introductory material on debate trees. I'm stll leaning towards using a simpler set of nodes to describe an argument, and am hoping to find a basic graphical step-by-step of an actual example to let me step through your model in my mind, and compare what you do with what I would otherwise do.
EDIT: Oh, my informal debate policy at this point is to engage if I otherwise have resources to do so and the person is being a scout about an interesting topic (using Galef's scout and soldier model). If they are being a soldier, I'll supply them information if that's in my interest. If they're being manipulative instead of a scout or soldier, I might still supply them information but I don't consider myself in an actual debate. At most, I'll spend a little time clarifying my position (to avoid intentional misinterpretations when that's feasible) then move on.
I partition behaviors into manipulative and truth-building, and partition truth-building behaviors into scout and soldier behaviors. By "behavior" I mean external behaviors that indicate internal processing indicative of scout or soldier mindset. I cannot actually debate someone being manipulative. Obviously, if my behavior is manipulative, then it doesn't qualify as genuine debate either.
EA is trying to affect the allocation of money. I suspect affecting the allocation of attention could be higher leverage and do more good. Put another way, ideas matter more than money, so the emphasis on directing donations isn't the highest leverage. Money does bring attention but isn't the most efficient way to get attention, let alone direct attention where you want, which is different than attracting attention. EA does effect attention, sometimes fairly intentionally and sometimes more as a side effect. But EA's goal is more about money than attention.
There may be other very important resources besides money or attention. Consideration should be given to what offers the most leverage to try to affect the allocation of. What is the best big picture goal or mix of goals for EA? Why emphasize money so much?
Being too busy to do something or engage with some idea means prioritizing some other activities or ideas. It's a standard excuse given to avoid giving any actual reason for deprioritizing the thing you're not doing, which prevents any truth-seeking about what priorities are correct.
I’ve seen the idea that getting other people to donate can be more effective than donating. It’s potentially higher leverage. What if you persuade ten people to donate? That action – that donation of some time and effort – does more good than donating money yourself (assuming you aren’t rich and couldn’t donate a ton).
Meta approaches often have higher leverage. They also have downsides and risks. If people only tried to persuade others, and no one wanted to donate, some problems would come up.
The right approach is to do some of the meta approach and some non-meta approach. All of one or the other is worse than a mix.
I forget where I read this, but I think so far I’m basically summarizing an existing EA idea. Now here’s my new thought:
There is a meta-meta-approach which is even higher leverage, and some of it should be done too. It is persuade people about how to do persuasion better – e.g. discussing rational debate methods. If you give ten people effective tips on how to debate rationally, then they can potentially each go persuade ten people to donate. The potential leverage is very high.
It's bad for all discussion topics to become inactive and ignored after around a day, like Reddit. The forum model, where topics go back to the top when anyone replies in them, is better for having longer, more serious or thoughtful discussions. Attention should be allocated to topics more based on writing activity instead of based on upvotes with a 1-2 day limit even for upvoted topics. Standard forums aren't a perfect system (e.g. they reward controversy) but they work better. I'm sure there's room for refinement of the forum model to help with problems if effort was put into a better system. But the basic current concept, that everything old gets ignored, is bad.
There's a big incentive to never write anything important or serious in comments, unless maybe a topic is new and popular (and it's hard to write important, serious things fast enough for that to work).
Interesting! I recently decided not to comment on a post that is not even a month old because there was no discussion about it anymore (and hadn't been even just a few days after it was posted) and I thought that my comment would be read by barely anyone. Not saying that this is generally a reasonable decision even in the current system, but sometimes it is even when one has something interesting to say... Maybe it's worth writing a longer post about this?
I don't know effective ways to cause EA to change, or, in the alternative, to get people to debate for the other side regarding my criticisms. And I don't want to play the social game of figuring out how to get the positive initial reactions, popularity and upvotes to draw more attention to my posts. I want an alternative truth-seeking game rather than to win more at the social game. But I don't know how to effectively say that to EA either.
Big agree! I call it the "frontpage time window". To mitigate the downsides, I suggested resurfacing curated posts on a spaced-repetition cycle to encourage long-term engagement, and/or breaking out a subset of topics into "tabs" so the forum effectively becomes something like a hybrid model of the current system and the more traditional format.
Why not just use a standard forum like discourse or phpbb? (I think the answer is that a more social media style website was created on purpose. There are things people prefer about it, though they mostly don't like to state them and may not consciously know them.) Also curation would be more about social status than truth. Having some central gatekeepers deciding what ideas are seen by many people is really problematic (having it decided with voting is also really problematic, and btw I think EA's weighted voting is worse than one vote per person).
There are both positives and negatives to being addicted to the EA Forum, so I'm not sure I'm against it entirely. Also, I think the current gatekeeper(s) (mainly Lizka atm, I think) have better judgment than the aggregated wisdom of accumulated karma, so I really appreciate the curation. Weighted upvotes are controversial, and I could change my mind, but for now I suspect the benefits outweigh the costs because I trust the judgment of high-karma-power users more on average. For an alternative way to filter posts, I think Zoe William's weekly summaries are excellent.
Ideally, it would be cool if EA karma could be replaced with a distributed trust network of sorts, such as one I alluded to on LW.
Truth seeking requires things like arguments, criticism and debate – and people who take responsibility for doing truth seeking. Without attempts to organize those things and make them happen following standardized methods, I don't think EA will be very effective. Substitutes like curation or voting are never going to get similar results.
I don't know of standardised methods that I think are likely to more reliably generate that, and I am worried about rushing standardisation prematurely. But this worry is far from insurmountable, and I'd be happy to hear suggestions for things that should be standardised. You have any? : )
Yes I've been posting suggestions here and no one has engaged seriously or offered any substantive criticism, e.g. I posted about my debate policy, my suggestion of an anti-misquoting norm, my explanation that rationality policies parallel the rule of law, a long list of sample policies, and some explanation of how people should take individual responsibility for their ideas and it's very hard to get anywhere with a group like EA when it's no one's job or responsibility to deal with criticism so everyone can individually just ignore it. There's also a continuity problem where basically each individual tends to see a random selection of what I post, which makes it hard to explain anything long, complicated and based on some different background knowledge than most people here believe or are familiar with. I agree that standardizing methods is a work in progress rather than finished, but it seems most people are not working on it, don't want to, and don't want to debate the matter. Do you want to try to have a serious discussion over time and try to reach some conclusions?
Hmm, not sure how that job would work, but if someone could be paid to fix all of EAs coordination problems, that'd probably be worth the money. It is the responsibility of anyone who wishes to assume that responsibility. And if they can, I really hope they do.
I'm a very slow reader, but do you wish to discuss (or debate it!) over a video call sometime? I'm eager to learn things from someone who's likely to have a different background on the questions I'm interested in. : )
I doubt that's going to work, but I can ask some questions about your proposal to check. My first question is whether you had in mind a public or private call.
Happy to either, but I'll stay off-camera if it's going to be recorded. Up to you, if you wish to prioritise it. : )
You didn't answer my question.
I'm confused. You asked whether I had in mind a public or private call, and I said I'd be fine with either. Which question are you referring to?
You didn't say which one you had in mind.
Mh, I had in mind both, and wanted to leave it up for interpretation. A public debate about something could be cool because I've never done that, but we'd need to know what we're supposed to disagree about first. Though, I primarily just wish to learn from you, since you have a different perspective, so a private call would be my second offer.
I consider calls inconvenient and inferior to writing, but I'll view it as a cultural difference and try to be flexible. So I'll do a recorded call, which I might put on YouTube, if we can agree on a time. Your listed times don't work for me but you seem to use this forum at later times. Would noon (US Pacific time) or later work for you? Nov 21?
Yes! That should work fine. That's 21:00 CET for me. See you then!
My email is emrik.asheim@gmail.com btw.
Ok what software? Is there an EA-related discord to meet and talk on?
My email is curi@curi.us
I certainly take you seriously, Elliot. However, I am strapped for time so I have not been able to review your video or more of your introductory material on debate trees. I'm stll leaning towards using a simpler set of nodes to describe an argument, and am hoping to find a basic graphical step-by-step of an actual example to let me step through your model in my mind, and compare what you do with what I would otherwise do.
EDIT: Oh, my informal debate policy at this point is to engage if I otherwise have resources to do so and the person is being a scout about an interesting topic (using Galef's scout and soldier model). If they are being a soldier, I'll supply them information if that's in my interest. If they're being manipulative instead of a scout or soldier, I might still supply them information but I don't consider myself in an actual debate. At most, I'll spend a little time clarifying my position (to avoid intentional misinterpretations when that's feasible) then move on. I partition behaviors into manipulative and truth-building, and partition truth-building behaviors into scout and soldier behaviors. By "behavior" I mean external behaviors that indicate internal processing indicative of scout or soldier mindset. I cannot actually debate someone being manipulative. Obviously, if my behavior is manipulative, then it doesn't qualify as genuine debate either.
EA is trying to affect the allocation of money. I suspect affecting the allocation of attention could be higher leverage and do more good. Put another way, ideas matter more than money, so the emphasis on directing donations isn't the highest leverage. Money does bring attention but isn't the most efficient way to get attention, let alone direct attention where you want, which is different than attracting attention. EA does effect attention, sometimes fairly intentionally and sometimes more as a side effect. But EA's goal is more about money than attention.
There may be other very important resources besides money or attention. Consideration should be given to what offers the most leverage to try to affect the allocation of. What is the best big picture goal or mix of goals for EA? Why emphasize money so much?
Being too busy to do something or engage with some idea means prioritizing some other activities or ideas. It's a standard excuse given to avoid giving any actual reason for deprioritizing the thing you're not doing, which prevents any truth-seeking about what priorities are correct.
I’ve seen the idea that getting other people to donate can be more effective than donating. It’s potentially higher leverage. What if you persuade ten people to donate? That action – that donation of some time and effort – does more good than donating money yourself (assuming you aren’t rich and couldn’t donate a ton).
Meta approaches often have higher leverage. They also have downsides and risks. If people only tried to persuade others, and no one wanted to donate, some problems would come up.
The right approach is to do some of the meta approach and some non-meta approach. All of one or the other is worse than a mix.
I forget where I read this, but I think so far I’m basically summarizing an existing EA idea. Now here’s my new thought:
There is a meta-meta-approach which is even higher leverage, and some of it should be done too. It is persuade people about how to do persuasion better – e.g. discussing rational debate methods. If you give ten people effective tips on how to debate rationally, then they can potentially each go persuade ten people to donate. The potential leverage is very high.