If you want to nuke trolling, use the Metafilter strategy: new accounts have to pay $5 (once). Troll too much, lose your account and pay $5 for a new one. Hurts a lot more than downvotes.
This will deter some (a lot?) of non-trolls from making new accounts. It will slow community growth. On the other hand, it will tighten the community and align interests. Casual users don’t contribute to Less Wrong’s mission: we need more FAI philanthropist/activists. Requiring a small donation will make it easier for casual users to make the leap to FAI philanthropist/activists, even if it makes it harder for lurkers to become casual users. And it will stop the trolling.
If you want to nuke trolling, use the Metafilter strategy: new accounts have to pay $5 (once).
I don’t know if I would have made my account here if I had to pay $5 to do so. I would pay $5 now to remain a member of the community- but I’ve already sunk a lot of time and energy into it. I mean, $5 is less cost to me than writing a new post for main!
I am deeply reluctant to endorse any strategy that might have turned me away as a newcomer.
What if you had to associate your account with a mobile phone number, by getting an activation code by text message? It still has the effect of requiring some real resource to make an account, but the first one is effectively free. There may be some concern about your number being sold to scammers.
If I encountered an unfamiliar blog or forum and wanted to leave a comment, I wouldn’t give my phone number to do so, even if it seemed quite interesting. Then I would probably leave the site.
Hard to say. So far, I’ve only given out my phone number to online services like gmail (woo 2 factor authentication!) or banks, but that’s because my email and bank accounts are more powerful than my phone number and because very few services ask for it. I think there’s a chance I wouldn’t give out my phone number, and I can’t clearly feel whether that chance is larger or smaller than my reluctance to pay $5. (Modeling myself from over a year ago is tough.)
This also runs into the trouble that instead of getting resources from users, you’re spending them on users- texting activation codes is cheap but not free.
Err really? I’d like to make some sort of bet on this—how many phone numbers you can receive texts from verses how many email addresses I can receive texts from by some deadline. Interested? You wouldn’t have to actually receive on them all of course, we’ll both use sampling to check.
You are, of course, correct. There’d be a bit of a delay—I was thinking of different email providers, not creating lots on one domain. And SIMs are sorta slow to turn over. But accumulating a pile of phone numbers for trolling would not be hard.
The “different email providers” thing is an interesting caveat, but how are you proposing to make use of that caveat in software? It’s not that it’s impossible on the face of it, but any software that wanted to make use of it would AFAICT have to have a painstakingly hand-crafted database of domain rules, so that you accept lots of gmail.com addresses but not lots of ciphergoth.org addresses.
It’s not like that in all countries. In Italy (unless the law has recently changed) you have to provide an identity document in order to activate a new SIM.
An even lower barrier would be a 100 captchas. That would accessible to almost everyone, and annoying to do repeatedly. Being a lower barrier though would mean it deters fewer trolls and doesn’t tighten the community as much.
I have attended one of the minicamps, and so have given money to SIAI, but as I believe the camp was run at a loss I wouldn’t count that as a donation.
I agree with this post, and am watching the creation of CFAR with interest.
An unintended side-effect: readers without credit/debit cards may find it harder to join the site. This disproportionately affects younger people, a demographic that may be more open to LW ideas.
Another unintended side-effect is that it may increase phyg pattern-matching. Now new recruits have to pay to join the site, and surely that money is being secretly funneled into EY’s bank account.
That said, I think that on balance this is a good policy proposal. I also think that the similar proposal using phone verification is plausible, and doesn’t run into the above two problems.
I don’t think anyone at SI agrees with you about Less Wrong’s mission. The site is supposed to be about rationality. There is hope (and track record) of the Less Wrong rationality community having helpful spinoffs for SI’s mission, but those benefits depend on it having its own independent life and mission. An open forum on rationality and a closed board for donors to a charity aren’t close substitutes for one another.
I think the percentage of “casual” users who participate on this site because they enjoy intelligent conversations on rationality-related topics while having no FAI agenda is non-negligible. I suspect that reinforcing the idea of equality between LW and FAI activism will make many of them leave. It may be a net negative even if LW’s mission is FAI activism as there are positive externalities of greater diversity of both discussion topics and participant opinions (less boredom, more new ideas, better critical scrutiny of ideas, less danger of community evaporative cooling, greater ability to attract new readers...)
Also, I don’t like the idea of LW’s mission being FAI activism. There is still written in the header: “A community blog devoted to refining the art of human rationality”, and I’d appreciate if I could continue believing that description. Of course I realise that the owners of the site are FAI enthusiasts, but that’s not true about the community as a whole. LW is a great rationality blog even without all its FAI/philanthropy stuff, not only for the texts already written, but also for the productive debating standards used here and a lot of intelligent people around. I would regret if I had to leave, which I would if LW turned to a solely FAI activist webpage.
I think the percentage of “casual” users who participate on this site because they enjoy intelligent conversations on rationality-related topics while having no FAI agenda is non-negligible.
I think they’re a majority, and that’s the problem. There’s a social norm that tolerates doing nothing about the most important problem humanity has ever faced.
It may be a net negative even if LW’s mission is FAI activism as there are positive externalities of greater diversity of both discussion topics and participant opinions (less boredom, more new ideas, better critical scrutiny of ideas, less danger of community evaporative cooling, greater ability to attract new readers...)
Charging a fee to comment is not the same as banning non-FAI topics. I agree that discussion of other topics can provide instrumental value to the SIAI.
Casual users don’t contribute to Less Wrong’s mission: we need more FAI philanthropist/activists.
The tagline is still “A community blog devoted to refining the art of human rationality”. If you want FAI and philanthropy, you should I suspect be asking for those specifically up front.
The tagline is still “A community blog devoted to refining the art of human rationality”. If you want FAI and philanthropy, you should I suspect be asking for those specifically up front.
The rationality discussion is a loss-leader, which brings smart, open-minded people into the shop. FAI activism is the high margin item LW needs to sell to remain profitable.
Right, but a tagline that knowingly omits important information about what you see as the actual mission will fairly obviously lead to (a) your time being wasted (b) their time being wasted. (And I’m not convinced a little logo to the side counts.) When you think the people who actually believe your tagline need to be made to go away, you may be doing something wrong.
The rationality discussion is a loss-leader, which brings smart, open-minded people into the shop. FAI activism is the high margin item LW needs to sell to remain profitable.
If that’s the case then LW is failing badly. There are a lot of people here like me who have become convinced by LW to be much more worried about existential risk at all but are not at all convinced that AI is a major segment of existential risk, and moreover even given that aren’t convinced that the solution is some notion of Friendliness in any useful sense. Moreover, this sort of phrasing makes the ideas about FAI sound dogmatic in a very worrying way. The Litany of Tarski seems relevant here. I want to believe that AGI is a likely existential risk threat if and only if AGI is a likely existential risk threat. If LW attracts or creates a lot of good rationalists and they find reasons why we should focus more on some other existential risk problem that’s a good thing.
If you want to nuke trolling, use the Metafilter strategy: new accounts have to pay $5 (once). Troll too much, lose your account and pay $5 for a new one. Hurts a lot more than downvotes.
It’s a good idea. Some variations, like associating accounts with mobile phone numbers, may slow good growth less. Maybe it would help to have multiple options to signal being a legitimate new user.
Casual users don’t contribute to Less Wrong’s mission: we need more FAI philanthropist/activists.
I don’t have a phone, and if I did I would refuse to give it out in case someone did something horrible like call me. I’m not the only phone-hater around; we overlap with phone-hater demographics a fair amount.
How would you feel about the $5 per account option?
Any other ideas on how someone could signal that the account they are creating is not yet another sock puppet or identity reset that you would be comfortable with? Maybe associating your account with your website?
I’m thinking the phone idea, if it is used at all, should be one of several options, so the user can choose one that works for them.
By strong default, I do not pay money for Internet intangibles, but $5 is low enough that I think we might see people buying accounts for their likely-valuable-commenter friends or something, so I’m not quite so opposed (but I think it would sharply slow community growth, and prevent people who we’d love to have around—like folks whose books get reviewed here—from dropping in to just say a few things).
I wouldn’t mind associating my website with my account—I already do, now that that’s an available field. But even fewer people have websites than phones.
Wouldn’t some kind of IP address thing suffice to rule out casually created socks?
This is an important point, we should be welcoming to people we talk about, and I’m not sure how that fits in to any scheme. Send out preemptive invitations when we talk about people? Who would keep on top of that?
I wouldn’t mind associating my website with my account—I already do, now that that’s an available field. But even fewer people have websites than phones.
Well, that was the result of me trying to find a mechanism that wouldn’t exclude you. But if we let people associate their account with a phone or a website, we include more people. It would be better to have more options to be more inclusive, if we can think of more specific options.
Wouldn’t some kind of IP address thing suffice to rule out casually created socks?
Yes, for certain values of casual. You can hide your IP address by going through proxies.
How would you feel about the $5 per account option?
That would show up on my credit card bill, which may cause certain inconveniences and I suspect we have people for whom that would cause a lot more than an inconvenience.
I think this is a VERY BAD IDEA. Charging $5 would have kept me out. It also keeps out everyone who doesn’t have a credit card, which includes basically every high school student.
What percentage of current posters do you estimate are FAI philanthopists/activists?
Can you give a couple specific examples of what distinguishes them from casual users? (donates to SI? works in a relevant field? volunteers for SI? etc)
Now that I think about that, neither poll asked takers how much they had donated for existential risk mitigation. (In case you’re wondering, the answer in my case would be “zero”.)
Can you give a couple specific examples of what distinguishes them from casual users? (donates to SI? works in a relevant field? volunteers for SI? etc)
If they donate a unit of caring to the SIAI (or perhaps the FHI), then I would lump them in the FAI philanthropist/activists category. There are some people who make contributions without donating money, such as Holden Karnofsky. But for the most part, the people who don’t give money are freeloaders.
They go to Less Wrong for entertainment, or for practical advice, or for social interaction. They don’t add value, and some take value away. They’ll direct the conversation to things that are fun to argue instead of things that that might save the world. Or they’ll write comments that show off their intellect instead of comments that raise the sanity waterline. I’m glad these people are gaining value from Less Wrong, but if we’re trying to save the world, they shouldn’t be our priority.
Edit: Changed the last paragraph to make it less mean.
It was a joke: Alicorn’s link in her own “quote” takes you to a comment that among other things says “(electric slide) I like to shake my butt, I like to make stuff up (electric slide) Is there published data? Maybe! Doesn’t matta! I’ll pull it out of my butt! (butt shake!)”
If you want to nuke trolling, use the Metafilter strategy: new accounts have to pay $5 (once). Troll too much, lose your account and pay $5 for a new one. Hurts a lot more than downvotes.
This will deter some (a lot?) of non-trolls from making new accounts. It will slow community growth. On the other hand, it will tighten the community and align interests. Casual users don’t contribute to Less Wrong’s mission: we need more FAI philanthropist/activists. Requiring a small donation will make it easier for casual users to make the leap to FAI philanthropist/activists, even if it makes it harder for lurkers to become casual users. And it will stop the trolling.
I don’t know if I would have made my account here if I had to pay $5 to do so. I would pay $5 now to remain a member of the community- but I’ve already sunk a lot of time and energy into it. I mean, $5 is less cost to me than writing a new post for main!
I am deeply reluctant to endorse any strategy that might have turned me away as a newcomer.
What if you had to associate your account with a mobile phone number, by getting an activation code by text message? It still has the effect of requiring some real resource to make an account, but the first one is effectively free. There may be some concern about your number being sold to scammers.
If I encountered an unfamiliar blog or forum and wanted to leave a comment, I wouldn’t give my phone number to do so, even if it seemed quite interesting. Then I would probably leave the site.
I suspect getting that to work in all countries would be a bit of a hassle.
Hard to say. So far, I’ve only given out my phone number to online services like gmail (woo 2 factor authentication!) or banks, but that’s because my email and bank accounts are more powerful than my phone number and because very few services ask for it. I think there’s a chance I wouldn’t give out my phone number, and I can’t clearly feel whether that chance is larger or smaller than my reluctance to pay $5. (Modeling myself from over a year ago is tough.)
This also runs into the trouble that instead of getting resources from users, you’re spending them on users- texting activation codes is cheap but not free.
I’d consider $5 but I would not have an account here if I had to buy a new phone in order to do so.
Do you know how many offers of free SIMs I get here in the UK? Really quite a lot. Phone numbers are as easy as email accounts.
Err really? I’d like to make some sort of bet on this—how many phone numbers you can receive texts from verses how many email addresses I can receive texts from by some deadline. Interested? You wouldn’t have to actually receive on them all of course, we’ll both use sampling to check.
You are, of course, correct. There’d be a bit of a delay—I was thinking of different email providers, not creating lots on one domain. And SIMs are sorta slow to turn over. But accumulating a pile of phone numbers for trolling would not be hard.
“A pile”, sure, but not millions.
The “different email providers” thing is an interesting caveat, but how are you proposing to make use of that caveat in software? It’s not that it’s impossible on the face of it, but any software that wanted to make use of it would AFAICT have to have a painstakingly hand-crafted database of domain rules, so that you accept lots of gmail.com addresses but not lots of ciphergoth.org addresses.
It’s not like that in all countries. In Italy (unless the law has recently changed) you have to provide an identity document in order to activate a new SIM.
An even lower barrier would be a 100 captchas. That would accessible to almost everyone, and annoying to do repeatedly. Being a lower barrier though would mean it deters fewer trolls and doesn’t tighten the community as much.
I can’t even solve single captchas and need to retry many times. You’d be seriously disadvantaging some people.
So I’m not the only one. (Is it my impression or did they use to be much easier until not long ago?)
Have you donated any money to the SIAI? Sorry for the personal question, but you did post a personal anecdote.
I have attended one of the minicamps, and so have given money to SIAI, but as I believe the camp was run at a loss I wouldn’t count that as a donation.
I agree with this post, and am watching the creation of CFAR with interest.
An unintended side-effect: readers without credit/debit cards may find it harder to join the site. This disproportionately affects younger people, a demographic that may be more open to LW ideas.
Another unintended side-effect is that it may increase phyg pattern-matching. Now new recruits have to pay to join the site, and surely that money is being secretly funneled into EY’s bank account.
That said, I think that on balance this is a good policy proposal. I also think that the similar proposal using phone verification is plausible, and doesn’t run into the above two problems.
Heck, there’s no pattern-matching about it. It will increase phyg.
No, it pattern matches Metafilter, like the top post said. Also SomethingAwful.
I never said anything about what it pattern-matches to.
(Edit: did you mean to reply to ParagonProtege’s comment?)
I don’t think anyone at SI agrees with you about Less Wrong’s mission. The site is supposed to be about rationality. There is hope (and track record) of the Less Wrong rationality community having helpful spinoffs for SI’s mission, but those benefits depend on it having its own independent life and mission. An open forum on rationality and a closed board for donors to a charity aren’t close substitutes for one another.
Who is “we”?
I think the percentage of “casual” users who participate on this site because they enjoy intelligent conversations on rationality-related topics while having no FAI agenda is non-negligible. I suspect that reinforcing the idea of equality between LW and FAI activism will make many of them leave. It may be a net negative even if LW’s mission is FAI activism as there are positive externalities of greater diversity of both discussion topics and participant opinions (less boredom, more new ideas, better critical scrutiny of ideas, less danger of community evaporative cooling, greater ability to attract new readers...)
Also, I don’t like the idea of LW’s mission being FAI activism. There is still written in the header: “A community blog devoted to refining the art of human rationality”, and I’d appreciate if I could continue believing that description. Of course I realise that the owners of the site are FAI enthusiasts, but that’s not true about the community as a whole. LW is a great rationality blog even without all its FAI/philanthropy stuff, not only for the texts already written, but also for the productive debating standards used here and a lot of intelligent people around. I would regret if I had to leave, which I would if LW turned to a solely FAI activist webpage.
I think they’re a majority, and that’s the problem. There’s a social norm that tolerates doing nothing about the most important problem humanity has ever faced.
Charging a fee to comment is not the same as banning non-FAI topics. I agree that discussion of other topics can provide instrumental value to the SIAI.
The tagline is still “A community blog devoted to refining the art of human rationality”. If you want FAI and philanthropy, you should I suspect be asking for those specifically up front.
The rationality discussion is a loss-leader, which brings smart, open-minded people into the shop. FAI activism is the high margin item LW needs to sell to remain profitable.
Right, but a tagline that knowingly omits important information about what you see as the actual mission will fairly obviously lead to (a) your time being wasted (b) their time being wasted. (And I’m not convinced a little logo to the side counts.) When you think the people who actually believe your tagline need to be made to go away, you may be doing something wrong.
If that’s the case then LW is failing badly. There are a lot of people here like me who have become convinced by LW to be much more worried about existential risk at all but are not at all convinced that AI is a major segment of existential risk, and moreover even given that aren’t convinced that the solution is some notion of Friendliness in any useful sense. Moreover, this sort of phrasing makes the ideas about FAI sound dogmatic in a very worrying way. The Litany of Tarski seems relevant here. I want to believe that AGI is a likely existential risk threat if and only if AGI is a likely existential risk threat. If LW attracts or creates a lot of good rationalists and they find reasons why we should focus more on some other existential risk problem that’s a good thing.
It’s a good idea. Some variations, like associating accounts with mobile phone numbers, may slow good growth less. Maybe it would help to have multiple options to signal being a legitimate new user.
I would like to see more x-risk philanthropists/activists, but I don’t want to make that a requirement for LW users. It would be good to have more users who want to be stronger because they have something to protect, rather than thinking rationality is shiny.
I don’t have a phone, and if I did I would refuse to give it out in case someone did something horrible like call me. I’m not the only phone-hater around; we overlap with phone-hater demographics a fair amount.
How would you feel about the $5 per account option?
Any other ideas on how someone could signal that the account they are creating is not yet another sock puppet or identity reset that you would be comfortable with? Maybe associating your account with your website?
I’m thinking the phone idea, if it is used at all, should be one of several options, so the user can choose one that works for them.
By strong default, I do not pay money for Internet intangibles, but $5 is low enough that I think we might see people buying accounts for their likely-valuable-commenter friends or something, so I’m not quite so opposed (but I think it would sharply slow community growth, and prevent people who we’d love to have around—like folks whose books get reviewed here—from dropping in to just say a few things).
I wouldn’t mind associating my website with my account—I already do, now that that’s an available field. But even fewer people have websites than phones.
Wouldn’t some kind of IP address thing suffice to rule out casually created socks?
This is an important point, we should be welcoming to people we talk about, and I’m not sure how that fits in to any scheme. Send out preemptive invitations when we talk about people? Who would keep on top of that?
Well, that was the result of me trying to find a mechanism that wouldn’t exclude you. But if we let people associate their account with a phone or a website, we include more people. It would be better to have more options to be more inclusive, if we can think of more specific options.
Yes, for certain values of casual. You can hide your IP address by going through proxies.
It would have false positives due to people sharing public IPs (but not computers) on workplace or campus networks.
And due to e.g. family members sharing IPs and computers.
That would show up on my credit card bill, which may cause certain inconveniences and I suspect we have people for whom that would cause a lot more than an inconvenience.
Strongly agree. There are also several objections raised on that comment.
I think this is a VERY BAD IDEA. Charging $5 would have kept me out. It also keeps out everyone who doesn’t have a credit card, which includes basically every high school student.
What percentage of current posters do you estimate are FAI philanthopists/activists?
Can you give a couple specific examples of what distinguishes them from casual users? (donates to SI? works in a relevant field? volunteers for SI? etc)
Now that I think about that, neither poll asked takers how much they had donated for existential risk mitigation. (In case you’re wondering, the answer in my case would be “zero”.)
I don’t want to pull a number out of my butt, but look at this way: our most recent open thread had 287 comments. Our summer fundraising thread had 17. We should probably have a survey.
If they donate a unit of caring to the SIAI (or perhaps the FHI), then I would lump them in the FAI philanthropist/activists category. There are some people who make contributions without donating money, such as Holden Karnofsky. But for the most part, the people who don’t give money are freeloaders.
They go to Less Wrong for entertainment, or for practical advice, or for social interaction. They don’t add value, and some take value away. They’ll direct the conversation to things that are fun to argue instead of things that that might save the world. Or they’ll write comments that show off their intellect instead of comments that raise the sanity waterline. I’m glad these people are gaining value from Less Wrong, but if we’re trying to save the world, they shouldn’t be our priority.
Edit: Changed the last paragraph to make it less mean.
I think you really should separate SIAI and LW here. I’d like to think that I’ve added value to LW in ways that don’t involve donating money to SIAI.
I’ll also note that it’s hard to freeload on social interaction.
Fixed your link.
Is this comment ironic? My link goes exactly where I intended: to a page that discusses the problems with attaching numerical estimates to everything.
It was a joke: Alicorn’s link in her own “quote” takes you to a comment that among other things says “(electric slide) I like to shake my butt, I like to make stuff up (electric slide) Is there published data? Maybe! Doesn’t matta! I’ll pull it out of my butt! (butt shake!)”
Where did you get that idea about “Less Wrong’s mission” from? Actually, when LW was created, discussing AI wasn’t even allowed.
Also discourages sockpuppetry