He did give a permission to restore the posts (I didn’t ask about comments), when I contacted him originally. There remains the issue of someone being technically able to restore these posts.
We have the technical ability, but it’s not easy.
We wouldn’t do it without Roko’s and Eliezer’s consent, and a lot of support for the idea.
(I wouldn’t expect Eliezer to consent to restoring the last couple of days of posts/comments, but we could restore everything else.)
It occurs to me that there is a call for someone unaffiliated to maintain a (scraped) backup of everything that is posted in order to prevent such losses in the future.
It’s doable. Are you now talking to the wrong person?
[ETA: Sorry—reading that back it was probably rude—I meant to say something closer to “It’s doable, but I still need Eliezer’s okay before I’ll do anything.”]
Okay granted. I also think this would be a good idea. Actually, I’d be against having an easy way to delete all contributions in general, it’s too easy to wreck things that way.
Actually, I’d be against having an easy way to delete all contributions in general, it’s too easy to wreck things that way.
Are you saying that the only person who should be conveniently able to remove other people’s contributions is you?
People’s comments are their own. It is unreasonable to leave them up if they choose not to. Fortunately, things that have been posted on the internet tend to be hard to destroy. Archives can be created and references made to material that has been removed (for example, see RationalWiki). This means that a blogger can not expect to be able to remove their words from the public record even though they can certainly stop publishing it themselves, removing their ongoing implied support of those words.
I actually do support keeping an archive of contributions and it would be convenient if LW had a way to easily restore lost content. It would have to be in a way that was either anonymized (“deleted user”?) or gave some clear indication that the post is by “past-Roko”, or “archived-Roko” rather than pretending that it is by the author himself, in the present tense. That is, it would acknowledge the futility of deleting information on the internet but maintain common courtesy to the author. There is no need to disempower the author ourselves by removing control over their own account when the very nature of the internet makes the deleting efforts futile anyway.
What is necessary is just that EY thinks about a way how to tell people why something had to be deleted without referring to what has been deleted in detail and why they should trust him on that. I see that freedom of speech has to end somewhere. Are we going to publish detailed blueprints for bio weapons? No. I just don’t see how EY wants to accomplish that as in the case of the Roko incident you cannot even talk about what has been deleted in a abstract way.
Convince me not to spread the original posts and comments as much as I can? How are you going to do that? I already posted another comment yesterday with the files that I deleted again after thinking about it. This is just too far and fuzzy for me to not play with the content in question without thinking twice.
What I mean is that I personally have no problem with censorship, if I can see why it had to be done.
I’ve been thinking about it by moving domains: Imagine that, instead of communicating by electromagnetic or sound waves, we encoded information into the DNA of custom microbes and exchanged them. Would there be any safe way to talk about even the specifics of why a certain bioweapon couldn’t be discussed?
I don’t think there is. At some point in weaponized conversation, there’s a binary choice between inflicting it on people and censoring it.
I wonder how the SIAI is going to resolve that problem if it caused nightmares inside the SIAI itself. Is EY going to solve it all by himself? If he was going to discuss it, then with whom, since he doesn’t know who’s strong enough beforehand? That’s just crazy. Time will end soon anyway, so why worry I guess. Bye.
Archives can be created and references made to material that has been removed (for example, see RationalWiki).
And not just the example you cite. RationalWiki has written an entire MediaWiki extension specifically for the purpose of saving snapshots of Web pages, as people trying to cover their tracks happens a lot on some sites we run regular news pages on (Conservapedia, Citizendium).
Memory holing gets people really annoyed, because it’s socially extremely rude. It’s the same problem as editing a post to make a commentator look foolish. There may be general good reasons for memory holing, but it must be done transparently—there is too much precedent for presuming bad faith unless otherwise proven.
A simple mechanism to put the saved evidence in the same place as the assertions concerning it, rather than out in the cloud, is not onerous in practice. Mind you, most of the disk load for RW is the images …
He did give a permission to restore the posts (I didn’t ask about comments), when I contacted him originally. There remains the issue of someone being technically able to restore these posts.
We have the technical ability, but it’s not easy. We wouldn’t do it without Roko’s and Eliezer’s consent, and a lot of support for the idea. (I wouldn’t expect Eliezer to consent to restoring the last couple of days of posts/comments, but we could restore everything else.)
It occurs to me that there is a call for someone unaffiliated to maintain a (scraped) backup of everything that is posted in order to prevent such losses in the future.
Surely it would be easy to restore just Roko’s posts, leaving his comments dead.
Also, if you don’t end up restoring them, it’s rather awkward that he’s in the top contributors list, with a practically dead link.
It’s doable. Are you now talking to the wrong person?
[ETA: Sorry—reading that back it was probably rude—I meant to say something closer to “It’s doable, but I still need Eliezer’s okay before I’ll do anything.”]
Okay granted. I also think this would be a good idea. Actually, I’d be against having an easy way to delete all contributions in general, it’s too easy to wreck things that way.
Are you saying that the only person who should be conveniently able to remove other people’s contributions is you?
People’s comments are their own. It is unreasonable to leave them up if they choose not to. Fortunately, things that have been posted on the internet tend to be hard to destroy. Archives can be created and references made to material that has been removed (for example, see RationalWiki). This means that a blogger can not expect to be able to remove their words from the public record even though they can certainly stop publishing it themselves, removing their ongoing implied support of those words.
I actually do support keeping an archive of contributions and it would be convenient if LW had a way to easily restore lost content. It would have to be in a way that was either anonymized (“deleted user”?) or gave some clear indication that the post is by “past-Roko”, or “archived-Roko” rather than pretending that it is by the author himself, in the present tense. That is, it would acknowledge the futility of deleting information on the internet but maintain common courtesy to the author. There is no need to disempower the author ourselves by removing control over their own account when the very nature of the internet makes the deleting efforts futile anyway.
What is necessary is just that EY thinks about a way how to tell people why something had to be deleted without referring to what has been deleted in detail and why they should trust him on that. I see that freedom of speech has to end somewhere. Are we going to publish detailed blueprints for bio weapons? No. I just don’t see how EY wants to accomplish that as in the case of the Roko incident you cannot even talk about what has been deleted in a abstract way.
Convince me not to spread the original posts and comments as much as I can? How are you going to do that? I already posted another comment yesterday with the files that I deleted again after thinking about it. This is just too far and fuzzy for me to not play with the content in question without thinking twice.
What I mean is that I personally have no problem with censorship, if I can see why it had to be done.
I’ve been thinking about it by moving domains: Imagine that, instead of communicating by electromagnetic or sound waves, we encoded information into the DNA of custom microbes and exchanged them. Would there be any safe way to talk about even the specifics of why a certain bioweapon couldn’t be discussed?
I don’t think there is. At some point in weaponized conversation, there’s a binary choice between inflicting it on people and censoring it.
Like the descoladores!
Hah, I didn’t realize someone else had already imagined it. Generalizing from multiple, independently-generated fictional evidence?
Awesome reply, thanks :-)
Didn’t know about this either, thanks Alicorn.
I wonder how the SIAI is going to resolve that problem if it caused nightmares inside the SIAI itself. Is EY going to solve it all by himself? If he was going to discuss it, then with whom, since he doesn’t know who’s strong enough beforehand? That’s just crazy. Time will end soon anyway, so why worry I guess. Bye.
And not just the example you cite. RationalWiki has written an entire MediaWiki extension specifically for the purpose of saving snapshots of Web pages, as people trying to cover their tracks happens a lot on some sites we run regular news pages on (Conservapedia, Citizendium).
Memory holing gets people really annoyed, because it’s socially extremely rude. It’s the same problem as editing a post to make a commentator look foolish. There may be general good reasons for memory holing, but it must be done transparently—there is too much precedent for presuming bad faith unless otherwise proven.
Seems like a heavy-weight solution. I’d just use http://webcitation.org/ (probably combined with my little program, archiver).
A simple mechanism to put the saved evidence in the same place as the assertions concerning it, rather than out in the cloud, is not onerous in practice. Mind you, most of the disk load for RW is the images …