I agree- the answer given in the FAQ isn’t a complete and valid response to the critics of the Singularity. But it was never meant to be; it was meant to be “short answers to common questions.” The SI’s longer responses to critics of the Singularity are mostly in peer-reviewed research; for example, in:
Of course, now I feel pretty bad for linking you to several hundred pages of arguments- which are often overlapping and repetitive, and which still don’t represent everything the SI has written on the subject (or even a majority, I think). If you have any specific criticisms of the SI’s ideas, it might be faster for you to post them here.
I feel even (slightly) worse, because none of the SI’s arguments have reached the level of evidence you’re comparing it to- eg Givewell’s analyses, and the disproofs of spoon-bending and new reading methods. But the SI isn’t capable of providing evidence that strong, because whether or not its claims were accurate, they would still be predictions of an abrupt future change, as opposed to claims about the efficacy of past actions. I do think that, where possible, the SI has tried to be very transparent- for example, in my opinion the SI’s last yearly progress report was around as thorough as Givewell’s last yearly progress reports- part 1,part 2, part 3, part 4.
(On a side note, it might interest you that Holden Karnofsky, co-founder of Givewell, also analyzed the SI and came to a very negative conclusion- posted here. His post is currently the most upvoted post of all time on LessWrong.)
Well, not technically in an internal publication, at least not an SI one—it was presented at the VIIIth European Conference on Computing and Philosophy, and published in their proceedings.
For an extended version of that paper in a peer-reviewed journal, see
(That one’s not actually an SI publication, but then neither was the Chalmers one. Nor is it specifically an answer to critics, but rather an elaboration of the SI argument.)
For an extended version of that paper in a peer-reviewed journal, see
Sotala, Kaj (2012) Advantages of Artificial Intelligences, Uploads, and Digital Minds. International Journal of Machine Consciousness 4 (1), 275-291. ( http://kajsotala.fi/Papers/DigitalAdvantages.pdf )
Oooh, thanks! I’d been wondering if there had been follow-up on that. I’ve found the first to be an interesting read, so I was curious to know if there were plans to expand on the topic and use it in movement-building strategies.
I agree- the answer given in the FAQ isn’t a complete and valid response to the critics of the Singularity. But it was never meant to be; it was meant to be “short answers to common questions.” The SI’s longer responses to critics of the Singularity are mostly in peer-reviewed research; for example, in:
Luke Muehlhauser and Anna Salamon (2012). Intelligence Explosion: Evidence and Import. In The Singularity Hypothesis, Springer. (http://singularity.org/files/IE-EI.pdf)
Carl Shulman and Nick Bostrom (2012). How Hard is Artificial Intellience?. In Journal of Consciousness Studies, Imprint Academic. (http://www.nickbostrom.com/aievolution.pdf)
Chalmers, D. (2010). “The Singularity: A Philosophical Analysis.” Journal of Consciousness Studies 17:7-65. (http://consc.net/papers/singularity.pdf)
Sotala, Kaj (2012) Advantages of Artificial Intelligences, Uploads, and Digital Minds. International Journal of Machine Consciousness 4 (1), 275-291. ( http://kajsotala.fi/Papers/DigitalAdvantages.pdf )
Of course, now I feel pretty bad for linking you to several hundred pages of arguments- which are often overlapping and repetitive, and which still don’t represent everything the SI has written on the subject (or even a majority, I think). If you have any specific criticisms of the SI’s ideas, it might be faster for you to post them here.
I feel even (slightly) worse, because none of the SI’s arguments have reached the level of evidence you’re comparing it to- eg Givewell’s analyses, and the disproofs of spoon-bending and new reading methods. But the SI isn’t capable of providing evidence that strong, because whether or not its claims were accurate, they would still be predictions of an abrupt future change, as opposed to claims about the efficacy of past actions. I do think that, where possible, the SI has tried to be very transparent- for example, in my opinion the SI’s last yearly progress report was around as thorough as Givewell’s last yearly progress reports- part 1, part 2, part 3, part 4.
(On a side note, it might interest you that Holden Karnofsky, co-founder of Givewell, also analyzed the SI and came to a very negative conclusion- posted here. His post is currently the most upvoted post of all time on LessWrong.)
Well, not technically in an internal publication, at least not an SI one—it was presented at the VIIIth European Conference on Computing and Philosophy, and published in their proceedings.
For an extended version of that paper in a peer-reviewed journal, see
Sotala, Kaj (2012) Advantages of Artificial Intelligences, Uploads, and Digital Minds. International Journal of Machine Consciousness 4 (1), 275-291. ( http://kajsotala.fi/Papers/DigitalAdvantages.pdf )
(That one’s not actually an SI publication, but then neither was the Chalmers one. Nor is it specifically an answer to critics, but rather an elaboration of the SI argument.)
Oooh, thanks! I’d been wondering if there had been follow-up on that. I’ve found the first to be an interesting read, so I was curious to know if there were plans to expand on the topic and use it in movement-building strategies.
Thanks, fixed!