Deleted. Don’t link to possible information hazards on Less Wrong without clear warning signs.
For example this is the link that was in the now deleted. I repeat it with the clear warning signs and observe that Charlie Stross (the linked to author) has updated his post so that it actually gives his analysis of the forbidden topic in question.
Warning: This link contains something defined as an Information Hazard by the lesswrong administrator. Do not follow it if this concerns you: Charlie Stross discusses Roko’s Basilisk. On a similar note: You just lost the game.
I wanted the link to be available if necessary just so that it makes sense to people when I say that Charlie Stross doesn’t know how decision theory works and his analysis is rubbish. Don’t even bother unless you are interested in categorizing various kinds of ignorant comments on the internet.
A standard warning would be good to have. It feels awkward trying to come up with a warning without knowing precisely what is to be warned about. In particular it isn’t clear whether you would have a strong preference (ie. outright insistence) that the warning doesn’t include specific detail that Roko’s Basilisk is involved. After all, some would reason that just mentioning the concept brings it to mind and itself causes potential harm (ie. You’ve already made them lose the game).
Unfortunately (or perhaps fortunately) all such “Information Hazard” warnings are not going to be particularly ambiguous because there just aren’t enough other things that are given that label.
For example this is the link that was in the now deleted. I repeat it with the clear warning signs and observe that Charlie Stross (the linked to author) has updated his post so that it actually gives his analysis of the forbidden topic in question.
Warning: This link contains something defined as an Information Hazard by the lesswrong administrator. Do not follow it if this concerns you: Charlie Stross discusses Roko’s Basilisk. On a similar note: You just lost the game.
I wanted the link to be available if necessary just so that it makes sense to people when I say that Charlie Stross doesn’t know how decision theory works and his analysis is rubbish. Don’t even bother unless you are interested in categorizing various kinds of ignorant comments on the internet.
It’ll do until we have a better standard warning.
A standard warning would be good to have. It feels awkward trying to come up with a warning without knowing precisely what is to be warned about. In particular it isn’t clear whether you would have a strong preference (ie. outright insistence) that the warning doesn’t include specific detail that Roko’s Basilisk is involved. After all, some would reason that just mentioning the concept brings it to mind and itself causes potential harm (ie. You’ve already made them lose the game).
Unfortunately (or perhaps fortunately) all such “Information Hazard” warnings are not going to be particularly ambiguous because there just aren’t enough other things that are given that label.