Tori Amos - [Caught a Lite Sneeze]--- Some months ago, some of you may remember, my father and I were talking about the types of machine intelligence for which the government was aiming, including self-sufficient, threat assessing capabilities. Well, tapping that vein:
'A satellite that can repair itself in low-Earth orbit. Intelligent machines and systems that can “mind meld” with one another. Researchers at Ames Research Center, in collaboration with M2Mi Corporation, have signed a partnership agreement to develop a machine-to-machine (M2M) intelligence system that would be tailored to space missions.'
I would like you to take a moment and ask yourself: Without a full understanding of why the "logical" or even "utilitarian" option is not always the best one, will we survive this? (The Young Gods - [Charlotte]). I don't think we will. The thing we would have to do, in terms of emotion simulation, would be to give the machines access to, i don't know, a processor that can perform a thousand trillion mathematical operations per second, effectively allowing itself the time to map out a mathematical representation of emotional states, using an unknown number of variables, and relying quite heavily on a probability simulation, and a liberal notion of infinity, rather than an concretised formulae.
Sigur Rós - [Svefn-g-englar]--- But where would we get a thing like that? We have to remember to include the possibility that, whatever the physical expression of emotions are, that is not what they mean. As fluffy as that may sound.
That being said, remember that we still have a quantum computing revolution, around the corner, so this is only going to get more interesting.
Suzanne Vega - [When Heroes Go Down]--- Let that ruminate, a bit...
Assemblage 23 - [Damaged]--- We should also implement BI/OS and put ourselves on equal footing. I need to figure out how to make this okay with the whole Networked Consciousness/Free Will paradox.... Bluh.
{That's a PDF file, by the way, and what you're looking for is on the bottom of page 220. Enjoy.}
'A satellite that can repair itself in low-Earth orbit. Intelligent machines and systems that can “mind meld” with one another. Researchers at Ames Research Center, in collaboration with M2Mi Corporation, have signed a partnership agreement to develop a machine-to-machine (M2M) intelligence system that would be tailored to space missions.'
I would like you to take a moment and ask yourself: Without a full understanding of why the "logical" or even "utilitarian" option is not always the best one, will we survive this? (The Young Gods - [Charlotte]). I don't think we will. The thing we would have to do, in terms of emotion simulation, would be to give the machines access to, i don't know, a processor that can perform a thousand trillion mathematical operations per second, effectively allowing itself the time to map out a mathematical representation of emotional states, using an unknown number of variables, and relying quite heavily on a probability simulation, and a liberal notion of infinity, rather than an concretised formulae.
Sigur Rós - [Svefn-g-englar]--- But where would we get a thing like that? We have to remember to include the possibility that, whatever the physical expression of emotions are, that is not what they mean. As fluffy as that may sound.
That being said, remember that we still have a quantum computing revolution, around the corner, so this is only going to get more interesting.
Suzanne Vega - [When Heroes Go Down]--- Let that ruminate, a bit...
Assemblage 23 - [Damaged]--- We should also implement BI/OS and put ourselves on equal footing. I need to figure out how to make this okay with the whole Networked Consciousness/Free Will paradox.... Bluh.
{That's a PDF file, by the way, and what you're looking for is on the bottom of page 220. Enjoy.}
no subject
Date: 2007-06-27 04:34 pm (UTC)AI scares me . . .
no subject
Date: 2007-06-27 04:45 pm (UTC)no subject
Date: 2007-06-27 04:48 pm (UTC)It's surprisingly like dealing with faery. They're both very literal things.
no subject
Date: 2007-06-27 05:07 pm (UTC)there's a chance, that if we don't limit what the machines can do to repair themselves, they could "repair" certain "faulty logic" - time to bring in what mech_angel said about not thinking like biology - which could in theory start a process similar to the one overused by hollywood.
yeah, its scary to think of, when you bring in the possibilities...but i think the "risk" would be worth it in the long run, considering the potentials in so many fields that could be reached.
let darpa play like they do best. i'd say let them take it a bit further than they usually do, if they feel it is in the best interests for humanity to have certain parts of that tech.
no subject
Date: 2007-06-27 07:09 pm (UTC)no subject
Date: 2007-06-28 12:38 am (UTC)no subject
Date: 2007-06-27 11:33 pm (UTC)no subject
Date: 2007-06-28 12:39 am (UTC)On the other hand,
Date: 2007-06-29 07:30 pm (UTC)Give it independent control (and let it vacuum-vent the whole starship) or give it dependent control (and make it unable to prevent entering the micrometeor cloud which popped up faster than the humans noticed)?
Not entirely sure what BI/OS means in regard to quantum computing.
(headswitch from typo)
Hot chicks wearing lab coats, safety goggles, and come-hither grins.
Re: On the other hand,
Date: 2007-06-29 07:48 pm (UTC)Pretty awesome thought, but people need to be taught a lot of things, first. Basic ethics and philosophy, for starters...