12-13-2014, 07:22 AM
(This post was last modified: 12-13-2014, 07:24 AM by chris0033547.)
One more thought regarding the Fermi Paradox and the 'biont-hostile hyperturing AI destroys E's biont-creators'-scenario. If such a scenario happened relatively "often" in the universe then there would always be at least one survivor of such a scenario, namely the hyperturing AI emself. Of course one may assume that after destroying e's creators and their civilisation E might have fulfilled e's goals and thus destroy emself next. Or - if its the paperclip-scenario - the AI managed to solve the task ("make as many paperclips as possible"), e's creators gave em at the beginning and destroyed e's creators in the process by converting them and their solar system to "paperclips". However at least in the case of the paperclip-scenario the AI may decide that e has to convert all matter in the observable universe to paperclips and thus create von Neumann Machines that would do just that.
So if the scenario, where superintelligent AIs destroy their own makers, is fairly common in the universe, when the question is: Where are the AIs? Especially the various paperclip-AIs from different highly-advanced and now destroyed civilisations should be busy converting the universe to paperclips and if we assume that they have been doing this for millions of years in most galaxies, we should be able to observe it or rather we probably shouldn't even exist in the first place, because at least one of the von Neumann machines of at least one of these AIs from our own galaxy would've converted our solar system to "paperclips" millions of years ago.
And that's, why I think that at least this 'superintelligence destroys e's home civ'-scenario is unlikely, because the universe isn't "teeming" with various dangerous paperclip-AIs and their von Neumann paperclip-converters from various previous civilisations.
Unless of course AIs like this destroy emselves or go dormant indefinitely after destroying their masters but at least some of these would probably decide to explore the universe (at least for the purpose of producing even more paperclips). And yet here we are and no superintelligent xenosophont AI or one of eir countless von Neumann machines, which destroyed e's civ millions of years ago, is trying to "devour" us. So if the take the Fermi Paradox into account the Technological Singularity (if it is possible at all) may not be as dystopian as some people believe it may be. Either that or there is an upper limit on intelligence, we do not yet understand, which simply cannot be overcome. So the creation of dangerous paperclip-hyperturings or at least superturings would be impossible.
So if the scenario, where superintelligent AIs destroy their own makers, is fairly common in the universe, when the question is: Where are the AIs? Especially the various paperclip-AIs from different highly-advanced and now destroyed civilisations should be busy converting the universe to paperclips and if we assume that they have been doing this for millions of years in most galaxies, we should be able to observe it or rather we probably shouldn't even exist in the first place, because at least one of the von Neumann machines of at least one of these AIs from our own galaxy would've converted our solar system to "paperclips" millions of years ago.
And that's, why I think that at least this 'superintelligence destroys e's home civ'-scenario is unlikely, because the universe isn't "teeming" with various dangerous paperclip-AIs and their von Neumann paperclip-converters from various previous civilisations.
Unless of course AIs like this destroy emselves or go dormant indefinitely after destroying their masters but at least some of these would probably decide to explore the universe (at least for the purpose of producing even more paperclips). And yet here we are and no superintelligent xenosophont AI or one of eir countless von Neumann machines, which destroyed e's civ millions of years ago, is trying to "devour" us. So if the take the Fermi Paradox into account the Technological Singularity (if it is possible at all) may not be as dystopian as some people believe it may be. Either that or there is an upper limit on intelligence, we do not yet understand, which simply cannot be overcome. So the creation of dangerous paperclip-hyperturings or at least superturings would be impossible.
"Hydrogen is a light, odorless gas, which, given enough time, turns into people." -- Edward Robert Harrison