5 ms·

One of my favorite sites. If you aren't familiar with it already, it will probably lead you to lesswrong.com. Also worth a good amount of time input.

by vitno 14y ago

tekromancr 14y ago
Lesswrong is a weird place. I can read an awesome, insightful article one moment, then something completely bat-shit insane the next.
- stcredzero 14y ago
  Example of batsheet?
  - AnthonyMouse 14y ago
    http://www.target.com/p/batman-sheet-set-twin/-/A-13999144 http://www.target.com/p/batman-sheet-set-twin/-/A-13999144 Sorry, it had to be done.
  - alsothat 14y ago
    A good example of LW going off the deep end is "Roko's basilisk" and Yudkowsky's response http://kruel.co/2013/01/12/rokos-basilisk-everything-you-need-to-know/ http://kruel.co/2013/01/12/rokos-basilisk-everything-you-nee... http://rationalwiki.org/wiki/Roko%27s_basilisk http://rationalwiki.org/wiki/Roko%27s_basilisk
    - Tichy 14y ago
      That sounds scary, but not crazy?
      - henrikschroder 14y ago
        The basilisk argument, in a nutshell, is this: At some point in the future, a near omniscient and omnipotent being, whose existence is not a certainty, will judge you and may punish you if it finds that the actions you take today are not pleasing, according to its undiscoverable morals, and the fact that you know about this increases the risk of you being punished. The parallells to religion should be obvious to anyone with the ability to see the bigger picture, but it's somehow reassuring that even the most rational of humans can transform into a religious nutjob and not realize it. The road to hell is paved with good intentions, etc. :-)
        Tichy 14y ago
        It sounds like Pascal's wager to me. I haven't really thought it through in much detail, though - for safety reasons...
    - gwern 14y ago
      That's not LW's problem, that's pretty much solely Yudkowsky's problem. Very few LW regulars agree with his response.
      - FeepingCreature 14y ago
        He himself doesn't even agree with his response. In the Reddit open thread on the topic he said he reacted rashly because the idea that somebody would be smart enough to come up with it and then stupid enough to post it publicly (his words, paraphrased) did not occur to him in advance. I think currently he's primarily deleting posts for it because once you've started that trend, there's not really that much to make up by stopping.
    - davidgerard 14y ago
      I started the RW article because LW readers distressed over the basilisk were emailing me (and other RW contributors) for help, because the matter is banned from discussion at LW. Even though the idea is utterly foolish, it would be inhuman to just tell them to go away; but individual emails don't scale. So, thinking that since secrecy hadn't worked the way out would be the way through, we quickly put that together. Then Warren Ellis and Charlie Stross publicised it and I've spent a coupla days getting it into fit condition for hits from the entire world ... Frankly, the article is still insane. (Look at the talk page for my girlfriend's reaction. She felt like throwing my laptop across the room for wasting a second of her time on this inanity.) This is because this shit is almost impossible to explain to normal humans. I have found it easier explaining Scientology to normal humans. Pretty much no-one on LessWrong endorses or believes in the basilisk ... they just advocate most of its component parts. LessWrong are generally a pretty good bunch, but this will be their Xenu. Unfair as that may be. And they did it to themselves. (gwern - if you're wondering why Dmytry can be quite such a dick at LW about this stuff, it's because he gets email from LW's distressed children too. He does his best to be a helpful, decent person.)
- jacoblyles 14y ago
  The problem with LW is that the old-timer's confidence is pegged at 100% for non-controversial stuff (like Bayes' Theorem) and highly controversial stuff (like the inevitability of unfriendly AI). It's essentially a cult of personality around Eliezer Yudkowsky, with little dissent. They use the non-controversial value-add stuff to draw new members into the cult. Eliezer has stated in interviews that the most important thing you can do is help him on his research if you can, and otherwise to make as much money as possible to give to his research, save a bit for your living expense. Now he has dozens of people that agree with him completely, a few hundred that mostly agree with him, and thousands that think he's a pretty swell guy and don't realize the shady depths of the philosophy he's pushing.
  - kiba 14y ago
    I thought it is well established in the lesswrong community that we shouldn't give money to Eliezer Yudkowsky until the Singularity Institute proves itself to be of value.
    - jacoblyles 14y ago
      The prevailing ideas (AKA Eliezer's ideas) do change. They used to be for AI research before they were against it. I am too bothered by the groupthink to hang around for long periods of time. I get enough of it IRL.
      - FeepingCreature 14y ago
        I'm pretty sure he's still for AI research, he just realized he'd been irresponsible about it. I'd much rather people change their mind as they gain greater understanding, as long as they document the reasoning.
        jacoblyles 14y ago
        It's just that it gives me the heebee-jeebees to see everybody change their mind together. Eliezer often cites a theorem from a mathematician that says that ideal rational agents can't disagree and applies it to normal conversation. In the LW circle this is used to bully people into conformity. No allowance is given for the time that it takes humans to think things through or negotiate on a set of priors based on their joint life experience. In practice at LW, it means that if you haven't conformed to a result that disturbs you within the course of a single conversation then you must be irrational. Because by Aumann's theorem rational agents will not remain in an undecided state. They really are a bunch of borg.

One of my favorite sites. If you aren't familiar with it already, it will probably lead you to lesswrong.com. Also worth a good amount of time input.

by vitno 14y ago

One of my favorite sites. If you aren't familiar with it already, it will probably lead you to lesswrong.com. Also worth a good amount of time input.

tekromancr 14y ago

Lesswrong is a weird place. I can read an awesome, insightful article one moment, then something completely bat-shit insane the next.

stcredzero 14y ago
Example of batsheet?
- AnthonyMouse 14y ago
  http://www.target.com/p/batman-sheet-set-twin/-/A-13999144 http://www.target.com/p/batman-sheet-set-twin/-/A-13999144 Sorry, it had to be done.
- alsothat 14y ago
  A good example of LW going off the deep end is "Roko's basilisk" and Yudkowsky's response http://kruel.co/2013/01/12/rokos-basilisk-everything-you-need-to-know/ http://kruel.co/2013/01/12/rokos-basilisk-everything-you-nee... http://rationalwiki.org/wiki/Roko%27s_basilisk http://rationalwiki.org/wiki/Roko%27s_basilisk
  - Tichy 14y ago
    That sounds scary, but not crazy?
    - henrikschroder 14y ago
      The basilisk argument, in a nutshell, is this: At some point in the future, a near omniscient and omnipotent being, whose existence is not a certainty, will judge you and may punish you if it finds that the actions you take today are not pleasing, according to its undiscoverable morals, and the fact that you know about this increases the risk of you being punished. The parallells to religion should be obvious to anyone with the ability to see the bigger picture, but it's somehow reassuring that even the most rational of humans can transform into a religious nutjob and not realize it. The road to hell is paved with good intentions, etc. :-)
      - Tichy 14y ago
        It sounds like Pascal's wager to me. I haven't really thought it through in much detail, though - for safety reasons...
  - gwern 14y ago
    That's not LW's problem, that's pretty much solely Yudkowsky's problem. Very few LW regulars agree with his response.
    - FeepingCreature 14y ago
      He himself doesn't even agree with his response. In the Reddit open thread on the topic he said he reacted rashly because the idea that somebody would be smart enough to come up with it and then stupid enough to post it publicly (his words, paraphrased) did not occur to him in advance. I think currently he's primarily deleting posts for it because once you've started that trend, there's not really that much to make up by stopping.
  - davidgerard 14y ago
    I started the RW article because LW readers distressed over the basilisk were emailing me (and other RW contributors) for help, because the matter is banned from discussion at LW. Even though the idea is utterly foolish, it would be inhuman to just tell them to go away; but individual emails don't scale. So, thinking that since secrecy hadn't worked the way out would be the way through, we quickly put that together. Then Warren Ellis and Charlie Stross publicised it and I've spent a coupla days getting it into fit condition for hits from the entire world ... Frankly, the article is still insane. (Look at the talk page for my girlfriend's reaction. She felt like throwing my laptop across the room for wasting a second of her time on this inanity.) This is because this shit is almost impossible to explain to normal humans. I have found it easier explaining Scientology to normal humans. Pretty much no-one on LessWrong endorses or believes in the basilisk ... they just advocate most of its component parts. LessWrong are generally a pretty good bunch, but this will be their Xenu. Unfair as that may be. And they did it to themselves. (gwern - if you're wondering why Dmytry can be quite such a dick at LW about this stuff, it's because he gets email from LW's distressed children too. He does his best to be a helpful, decent person.)
jacoblyles 14y ago
The problem with LW is that the old-timer's confidence is pegged at 100% for non-controversial stuff (like Bayes' Theorem) and highly controversial stuff (like the inevitability of unfriendly AI). It's essentially a cult of personality around Eliezer Yudkowsky, with little dissent. They use the non-controversial value-add stuff to draw new members into the cult. Eliezer has stated in interviews that the most important thing you can do is help him on his research if you can, and otherwise to make as much money as possible to give to his research, save a bit for your living expense. Now he has dozens of people that agree with him completely, a few hundred that mostly agree with him, and thousands that think he's a pretty swell guy and don't realize the shady depths of the philosophy he's pushing.
- kiba 14y ago
  I thought it is well established in the lesswrong community that we shouldn't give money to Eliezer Yudkowsky until the Singularity Institute proves itself to be of value.
  - jacoblyles 14y ago
    The prevailing ideas (AKA Eliezer's ideas) do change. They used to be for AI research before they were against it. I am too bothered by the groupthink to hang around for long periods of time. I get enough of it IRL.
    - FeepingCreature 14y ago
      I'm pretty sure he's still for AI research, he just realized he'd been irresponsible about it. I'd much rather people change their mind as they gain greater understanding, as long as they document the reasoning.
      - jacoblyles 14y ago
        It's just that it gives me the heebee-jeebees to see everybody change their mind together. Eliezer often cites a theorem from a mathematician that says that ideal rational agents can't disagree and applies it to normal conversation. In the LW circle this is used to bully people into conformity. No allowance is given for the time that it takes humans to think things through or negotiate on a set of priors based on their joint life experience. In practice at LW, it means that if you haven't conformed to a result that disturbs you within the course of a single conversation then you must be irrational. Because by Aumann's theorem rational agents will not remain in an undecided state. They really are a bunch of borg.