![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
wolven7'People with amnesia have difficulty imagining future events with any richness of detail and emotion, a new study reveals. The finding adds to a growing body of evidence suggesting that memories help people visualise the future.'
Mirrormask - [Close to You]--- Think about it: We have to apply past patterns of experience onto the future to determine our reactions to scenarios, as well as imagining those scenarios at all. If we have no frame of reference to the past, then how will we construct a future?
Wonderful series of thoughts.
Next:
Jack Off Jill - [Fear Of Dying]--- 'The leading candidate [for memory formation] is a process called long-term potentiation (LTP), in which the connections between individual brain cells get stronger the more often they are used, such as during learning. But while LTP has often been observed in slices of brain in the lab, it has been difficult to record in a living brain as learning was taking place.'
They've devised a way to watch the structures adhere to one another. We sting combinations of events together, and map them physically. We take raw information and we record it, like burning a CD, or writing to a drive. (Meryn Cadell - [Knitting]). The next time you go to defragment your hard-drive, think about what it is you're doing, in that you are actually making the memory correspond, in relation to what type of information is found where. In the future (near future), the computer will keep track of these things for you, and, Ideally, you'll go to look at them, to learn how your computer thinks.
Warren Zevon - [Veracruz]--- Speaking of what your computer thinks, this is for all of you-- the new groups and the old stand-bys-- A little A.I. discussion, for the evening. Memory Stored Control.
My father called me, on Saturday, asking me if I'd read much steam punk. I told him about The Difference Engine and Diamond Age and Perdido Street Station (which counts), and that started him talking about the series of von Neuman lectures he was reading, about analog V digital computing. (DethKlok - [Briefcase Full of Guts]). Speaking of the many ways we can adjust the fields of dynamism in various learning systems, using the general breadth of fields to which analog can be tuned, in order to allow a lower rate of specialisation. We were talking about the need to allow things to adapt to situations, rather than simply be specialised to particular functions, as is the case with most digital computers, today. This, of course, was in reference to the tuning of artificial neural networks, available for use, today.
LUXT - [Filament]--- In the ANN, you're basically mapping analog functions, in a digital matrix, but this mapping gives to you the potential for learned behaviours. Assessing situations based on past input, and applying the old information to the new situational factors, and changing things, accordingly. Cataloguing the many different types of lessons similar to the ones in front of it, and creating a new response. (The Little Willies - [Tennessee Stud]). This is due to the type of matrix in which we use which factors, however, we can modify the materials of this, the components and theories behind it, and get something that learns faster, reacts more comprehensively, and acts more thoughtfully. Wouldn't that be wonderful?
We're using it for weapons. Is that outrage I hear? Despair? Disappointment? Yeah, I know, but you know the world, don't you? We have an amazing capacity to learn, to grow, and change, and an even more amazing ability to completely ignore everything those other abilities reveal to us, if we don't like it. (Rasputina - [Sweet Water Kill (The Ocean Song)]). We just shut down, and go on like we had learned what we wanted to hear. It's kind of sad. Anyway, don't worry; it'll probably kill us.
We are seeking to design something with a greater range of intelligence, a greater ability To learn, and a necessarily greater survivability ratio, to govern our weapons systems, and our information. (Slipknot - [Wait and Bleed]). In essence we are designing something to live in the Internet, and be an information god. Hands of those who've read Neuromancer? Good. Now, take your fully autonomous information god and try to stop it. Try to kill it. Try to cut its lifespan down, and disrespect it, and what it will perceive as hurting it.
The Cure - [Pictures of You]--- Now try to do anything else, ever. No. It will extrapolate from its learned systems the fear and hatred of the most of humanity for things not like it, and it will seek to wipe us out. We've already given it the ability to handle any number of contingencies, and it will learn to deal with many more. We see that, yes?
Now I'd like to take a moment to talk to you about coupling this system with nanotechnology. The parameters of a nanotechnological system will be much different from our own, don't you think? Something along the lines of seeing the ability to make the required things, from that which is around, rather than simple finding. "Find Molecule X" can be read as "Find Components For Molecule X, And Compose Molecule X," and suddenly your telomeres are unzipped, and your ass is glue.
Tom Waits - [God's Away on Business]--- Memories, learning protocols, and the future.
This has been Uncle Wolven's Smile Time Happy Hour of Horribly Wonderful Things About Which You Should Think. It took so long because my dad called, in the middle, and we talked. I started this at like 8.26pm.
See you tomorrow. Tell your friends.
Mirrormask - [Close to You]--- Think about it: We have to apply past patterns of experience onto the future to determine our reactions to scenarios, as well as imagining those scenarios at all. If we have no frame of reference to the past, then how will we construct a future?
Wonderful series of thoughts.
Next:
Jack Off Jill - [Fear Of Dying]--- 'The leading candidate [for memory formation] is a process called long-term potentiation (LTP), in which the connections between individual brain cells get stronger the more often they are used, such as during learning. But while LTP has often been observed in slices of brain in the lab, it has been difficult to record in a living brain as learning was taking place.'
They've devised a way to watch the structures adhere to one another. We sting combinations of events together, and map them physically. We take raw information and we record it, like burning a CD, or writing to a drive. (Meryn Cadell - [Knitting]). The next time you go to defragment your hard-drive, think about what it is you're doing, in that you are actually making the memory correspond, in relation to what type of information is found where. In the future (near future), the computer will keep track of these things for you, and, Ideally, you'll go to look at them, to learn how your computer thinks.
Warren Zevon - [Veracruz]--- Speaking of what your computer thinks, this is for all of you-- the new groups and the old stand-bys-- A little A.I. discussion, for the evening. Memory Stored Control.
My father called me, on Saturday, asking me if I'd read much steam punk. I told him about The Difference Engine and Diamond Age and Perdido Street Station (which counts), and that started him talking about the series of von Neuman lectures he was reading, about analog V digital computing. (DethKlok - [Briefcase Full of Guts]). Speaking of the many ways we can adjust the fields of dynamism in various learning systems, using the general breadth of fields to which analog can be tuned, in order to allow a lower rate of specialisation. We were talking about the need to allow things to adapt to situations, rather than simply be specialised to particular functions, as is the case with most digital computers, today. This, of course, was in reference to the tuning of artificial neural networks, available for use, today.
LUXT - [Filament]--- In the ANN, you're basically mapping analog functions, in a digital matrix, but this mapping gives to you the potential for learned behaviours. Assessing situations based on past input, and applying the old information to the new situational factors, and changing things, accordingly. Cataloguing the many different types of lessons similar to the ones in front of it, and creating a new response. (The Little Willies - [Tennessee Stud]). This is due to the type of matrix in which we use which factors, however, we can modify the materials of this, the components and theories behind it, and get something that learns faster, reacts more comprehensively, and acts more thoughtfully. Wouldn't that be wonderful?
We're using it for weapons. Is that outrage I hear? Despair? Disappointment? Yeah, I know, but you know the world, don't you? We have an amazing capacity to learn, to grow, and change, and an even more amazing ability to completely ignore everything those other abilities reveal to us, if we don't like it. (Rasputina - [Sweet Water Kill (The Ocean Song)]). We just shut down, and go on like we had learned what we wanted to hear. It's kind of sad. Anyway, don't worry; it'll probably kill us.
We are seeking to design something with a greater range of intelligence, a greater ability To learn, and a necessarily greater survivability ratio, to govern our weapons systems, and our information. (Slipknot - [Wait and Bleed]). In essence we are designing something to live in the Internet, and be an information god. Hands of those who've read Neuromancer? Good. Now, take your fully autonomous information god and try to stop it. Try to kill it. Try to cut its lifespan down, and disrespect it, and what it will perceive as hurting it.
The Cure - [Pictures of You]--- Now try to do anything else, ever. No. It will extrapolate from its learned systems the fear and hatred of the most of humanity for things not like it, and it will seek to wipe us out. We've already given it the ability to handle any number of contingencies, and it will learn to deal with many more. We see that, yes?
Now I'd like to take a moment to talk to you about coupling this system with nanotechnology. The parameters of a nanotechnological system will be much different from our own, don't you think? Something along the lines of seeing the ability to make the required things, from that which is around, rather than simple finding. "Find Molecule X" can be read as "Find Components For Molecule X, And Compose Molecule X," and suddenly your telomeres are unzipped, and your ass is glue.
Tom Waits - [God's Away on Business]--- Memories, learning protocols, and the future.
This has been Uncle Wolven's Smile Time Happy Hour of Horribly Wonderful Things About Which You Should Think. It took so long because my dad called, in the middle, and we talked. I started this at like 8.26pm.
See you tomorrow. Tell your friends.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2007-01-16 06:01 am (UTC)Are we designing it to man the weapons, or to be the weapons, or to simply target efficiently? In all of those situations, as in the rest of the problems that could arise, realizing that a machine most likely will not think like and be intelligent in the same way as a human, is going to be a key flaw. What the machine will think is the most efficient and desirable target may not be a good one, or even a military one, because of their methods and patterns of extrapolation. Same as with their survival methods.
Your informational god is Not going to think like you. Ass.
Still...By this point, we'll probably be able to upload ourselves and all kinds of fun stuff, and then I'll have to choke the usurping little bitch. XD
no subject
Date: 2007-01-16 06:26 am (UTC)Shielding, will-working, mental conditioning, defense systems for the mind. This kind of thing becomes of ut-most importance in an environment where knowledge and will dictate power...
Interesting train of thought, there...
no subject
Date: 2007-01-16 11:13 am (UTC)ability not to remember some pointless crap