Donate Bitcoin

Donate Paypal


PeakOil is You

PeakOil is You

Techno-Future ramblings (psst Omnitir)

What's on your mind?
General interest discussions, not necessarily related to depletion.

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Omnitir » Wed 30 May 2007, 20:50:47

$this->bbcode_second_pass_quote('UFCjunkie', 'w')ithout oil we won't devolop more new techs we wont even be able to produce materia for the experiments. Or what is yor plan with that? How are we going to keep the technology growt in a world without oil?

Firstly, there will be oil. It will just be really expensive. Secondly, if tech keeps developing for only two more decades and we get MNT, we won't even need fossil fuels at all, since we then have the ability to produce material at the molecular level based on the raw materials (mostly carbon).

Yeah I know, star trek dream land, right? We'll see.

$this->bbcode_second_pass_quote('UFCjunkie', '
')$this->bbcode_second_pass_quote('Omnitir', ' ')If the singularity happens, and we end up with greatly increasing the intelligence on the planet, such a super intelligence will be able to solve our problems.

I believe that such intelligance would wipe out our race in one of the steps to solve this planets problems.

That's that common conclusion that people come to when they anthropomorphise an artificial intelligence. If this intelligence was like a human mind, than it probably would destroy all humans. But the human mind is only one small area of all possible minds. It won't necessarily think like we do. It depends how we build it. If it's built the right way, it could naturally desire to help people.

Of course wiping people out isn't the only way to get people to stop destorying the earth :)


You know, in a strange way, I hope you are right. I would much rather see civilization collapse back down to a pseudo sustainable hunter gatherer society than see it blow itself up on the road to further development. But I think this collapse scenario is actually the most dream land fantasy of all. No way are those with all the power in this world going to sit by and let that power go. They are going to do everything they can to keep any semblance of the status quo.

Some people act like once oil peaks (with apparently was several years ago), we are just going to have a few rights, some starvation, and back to hunter gatherers we go.

Meanwhile in the real world, those with power are waging resource wars, all the while pumping massive investment dollars into R&D to keep their armies on the cutting edge, and to keep their customers with the latest gizmo's.

Yeah, energy is intimately related to progress. But energy isn't going to disappear overnight. It's going to get expensive, it's going to get more violent to attain, and it's going to get dirtier. But energy, the most valuable commodity in the world, is not going to go away in a hurry. And as long as we have energy, we have progress.

If we make it through these turbulent times ahead, than we will achieve new and better ways of doing things. If we don't make it through, well it will likely be because we destroy ourselves. It seems most fairy tail to me that we could somehow get through these issues without either destroying ourselves or achieving technological sustainability.
"Mother Nature is a psychopathic bitch, and she is out to get you. You have to adapt, change or die." - Tihamer Toth-Fejel, nanotech researcher/engineer.
User avatar
Omnitir
Tar Sands
Tar Sands
 
Posts: 894
Joined: Sat 02 Apr 2005, 04:00:00
Location: Down Under

Re: Techno-Future ramblings (psst Omnitir)

Unread postby TheDude » Thu 31 May 2007, 00:45:01

Hmm, lots of things to address.

$this->bbcode_second_pass_quote('Omnitir', 'T')hat's that common conclusion that people come to when they anthropomorphise an artificial intelligence. If this intelligence was like a human mind, than it probably would destroy all humans. But the human mind is only one small area of all possible minds. It won't necessarily think like we do. It depends how we build it. If it's built the right way, it could naturally desire to help people.


Mmmm, ever read Jack Williamson's story With Folded Hands? Robots protect humanity - to a fault. Won't let us turn on burners - too dangerous. Perhaps a reductio ad absurdum but it gets the point across.

Will read up a bit more, listening to Kurzweil on YouTube right now. Haven't heard him point out how many hundreds of barrels of oil go into a measly microchip yet.

$this->bbcode_second_pass_quote('', 'Y')ou know, in a strange way, I hope you are right. I would much rather see civilization collapse back down to a pseudo sustainable hunter gatherer society than see it blow itself up on the road to further development. But I think this collapse scenario is actually the most dream land fantasy of all. No way are those with all the power in this world going to sit by and let that power go. They are going to do everything they can to keep any semblance of the status quo.


I asked this on another topic - where are they going to do that? I can see the Inner Party cordoning off Illinois or something equally extreme; there will still be an awful lot of really pissed off hungry people wanting to slit their throats. I think they're going to be a lot more keen on manafacturing 7.62mm cartridges than pursuing AI research. Anybody that'd let the rest of the world starve while they go on their merry way is a bit amoral. Hey, sounds like what we've got going right now...

$this->bbcode_second_pass_quote('', 'S')ome people act like once oil peaks (with apparently was several years ago), we are just going to have a few rights, some starvation, and back to hunter gatherers we go.


I'd say the more common outlook here is pre-Industrial. Some people on this forum like gg3 are serious about the concept of Powerdown, retaining our society but minimizing how much we use, and I'm for that - hope the rest of the US likes the idea.

You shouldn't get too hyperbolic about the Singularity, Om, otherwise people will think of this quote:

$this->bbcode_second_pass_quote('', 'K')en MacLeod describes the Singularity as “the Rapture for nerds” in his 1998 novel The Cassini Division.


Please keep posting though, I at least appreciate what you have to say. Ignore detractors like pstarr, I'm guessing he's got an unhealthy obsession with his Post Count. I know the feeling, I'm always posting short snide comments myself, to keep up the yucks here and also because I want to make sure people vote Cthulhu...

Another good story:

$this->bbcode_second_pass_quote('', 'T')he danger from that storage device and manipulator of data that we call the computer may have been summed up best in a 1954short-short story by Fredric Brown titled "Answer." Is there a critical mass for consciousness? A computer with as many junctions as the brain has synapses, scientists tell us, would be as big as a city or a city block, or something like that. Does it take that many junctions to create consciousness? In Brown's story computers across the galaxy are linked together and are asked the question that has been concerning humanity for as long as humanity has been around, "Is there a God?" And the computer answers, as it seals shut the switch with a lightning bolt, "There is now."
Cogito, ergo non satis bibivi
And let me tell you something: I dig your work.
User avatar
TheDude
Expert
Expert
 
Posts: 4896
Joined: Thu 06 Apr 2006, 03:00:00
Location: 3 miles NW of Champoeg, Republic of Cascadia
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby johnmarkos » Thu 31 May 2007, 01:14:00

$this->bbcode_second_pass_quote('Omnitir', 'I')t won't necessarily think like we do. It depends how we build it. If it's built the right way, it could naturally desire to help people.

Yeah, I was reading recently that our brains give us the same biochemical rewards for altruistic behavior that they do for food and sex. If we can reproduce a similar reward system in bots, maybe they won't kill us all. They'll feel good when they help us. So all we have to do is make the AIs altruism fetishists, just like humans, maybe more so. Make them extra good and extra smart.

Of course, once they start building better versions of themselves, they could rip out those silly three laws and come up with their own morality. Nonetheless, if they start out with a desire to be good hard wired in their little positronic brains, maybe we have a chance of that morality influencing the designs they make, too.
User avatar
johnmarkos
Tar Sands
Tar Sands
 
Posts: 866
Joined: Wed 19 May 2004, 03:00:00
Location: San Francisco, California
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby TheDude » Thu 31 May 2007, 02:52:28

One would certainly hope they're not working on a superpowerful AI version of Dick Cheney. Which is related to my point regarding the Haves sequestering themselves from us Have Nots. Bit of a philosophical minefield here.
I found a post from a few years back on Peak Oil and the Omega Point.
Cogito, ergo non satis bibivi
And let me tell you something: I dig your work.
User avatar
TheDude
Expert
Expert
 
Posts: 4896
Joined: Thu 06 Apr 2006, 03:00:00
Location: 3 miles NW of Champoeg, Republic of Cascadia

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Omnitir » Thu 31 May 2007, 04:47:13

$this->bbcode_second_pass_quote('TheDude', '
')Mmmm, ever read Jack Williamson's story With Folded Hands? Robots protect humanity - to a fault. Won't let us turn on burners - too dangerous. Perhaps a reductio ad absurdum but it gets the point across.

No, but I've read short stories with similar plots. Azimov wrote a few like this. It seems like a logical issue, until you consider the super intelligence factor. Considering an AI would rapidly become smarter than humans (because it can rewrite it's own code), than it would surely be able to work out that protecting us to a fault would cause us more harm than good.

$this->bbcode_second_pass_quote('TheDude', '
')Will read up a bit more, listening to Kurzweil on YouTube right now. Haven't heard him point out how many hundreds of barrels of oil go into a measly microchip yet.

Kurzweil is clear about his position on our energy future. It is the same position as many of the worlds leading scientists, only Kurzweil is of course more optimistic than most.

There is more energy in sunlight striking the Earth everyday than we could ever use, and our energy future, according to these guys, will be harnessing that power thanks to the exponentially increasing cost performance of solar power technology - made possible through nanotech.

$this->bbcode_second_pass_quote('Kurzweil', '"')Even though our energy needs are projected to triple within 20 years," he wrote recently, "we'll capture that .0003 of the sunlight needed to meet all of our energy needs with no use of fossil fuels, by using extremely inexpensive, highly efficient, lightweight, nano-engineered solar panels. From there, the energy will be stored in safe, highly distributed fuel cells."

"Solar power," he insists "is now providing one part in a thousand of our energy needs but that percentage is doubling every two years, which means multiplying by a thousand in 20 years. Almost all of the discussions I've seen about energy fail to consider the ability of future nanotechnology-based solutions to solve this problem. This development will be motivated not just by concern for the environment, but by the $2 trillion we spend annually on energy. This is already a major area of venture funding."


$this->bbcode_second_pass_quote('TheDude', 'Y')ou shouldn't get too hyperbolic about the Singularity, Om, otherwise people will think of this quote:

$this->bbcode_second_pass_quote('', 'K')en MacLeod describes the Singularity as “the Rapture for nerds” in his 1998 novel The Cassini Division.


lol, yeah, I usually avoid talking about it, but I'm indulging due to the thread topic.

However, I'm happy to talk about nanotech. It's got the potential to solve many issues, yet people dismiss it because it's at least a decade away. Yet then they turn around and talk about the great die-off that will unfold over the next 100 years...
Last edited by Omnitir on Thu 31 May 2007, 05:17:30, edited 1 time in total.
"Mother Nature is a psychopathic bitch, and she is out to get you. You have to adapt, change or die." - Tihamer Toth-Fejel, nanotech researcher/engineer.
User avatar
Omnitir
Tar Sands
Tar Sands
 
Posts: 894
Joined: Sat 02 Apr 2005, 04:00:00
Location: Down Under
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Omnitir » Thu 31 May 2007, 05:15:55

$this->bbcode_second_pass_quote('johnmarkos', '')$this->bbcode_second_pass_quote('Omnitir', 'I')t won't necessarily think like we do. It depends how we build it. If it's built the right way, it could naturally desire to help people.

Yeah, I was reading recently that our brains give us the same biochemical rewards for altruistic behavior that they do for food and sex. If we can reproduce a similar reward system in bots, maybe they won't kill us all. They'll feel good when they help us. So all we have to do is make the AIs altruism fetishists, just like humans, maybe more so. Make them extra good and extra smart.

Of course, once they start building better versions of themselves, they could rip out those silly three laws and come up with their own morality. Nonetheless, if they start out with a desire to be good hard wired in their little positronic brains, maybe we have a chance of that morality influencing the designs they make, too.


I agree with your first paragraph but have doubts about the second. You're right about how we need to design them to be inherently good from the beginning. But I think the Asimov "three laws safe" concept has distorted common concepts of AI somewhat. As you say, the three laws are silly. Asimov made them up for dramatic purposes and showed how complex it is to make an AI safe - the three laws don't work. But the notion that an AI will re-write it's morals I think is also misguided. As it rewrites it's code to become more intelligent, meaning to calculate quicker, why would it change it's basic purpose?

I think if an AI was designed from the beginning to be a certain way, then that is what it is, and no matter how advanced it becomes, as it rewrite it's code to increase it's performance, it will not spontaneously change it's basic premise from which it was originally designed.

So for example, if the military build the first strong AI to intelligently take out hostiles, then as this AI rewrites it's code to increase it's performance and efficiency, it will still desire to take out hostiles. It is most unlikely to spontaneously decide to change it's super goals to bring peace on earth.

Or alternatively, if a university builds the first AI to intelligently solve social problems to make life better for humans, once this AI starts to optimise and advance it's code it seems most unlikely that it will spontaneously decide to change it's fundamental supergoal of making human life better to making it's own life better.

It seems likely that the worlds first super AI will be the result of it's design. But not because of a bunch of laws telling what to do or not do - it will rewrite those to increase it's performance. But it will remain the result of it's original design because that is it's purpose, it's moral base.

Let's just hope that the military doesn't build Skynet before a university or business builds Friendlynet. 8O
"Mother Nature is a psychopathic bitch, and she is out to get you. You have to adapt, change or die." - Tihamer Toth-Fejel, nanotech researcher/engineer.
User avatar
Omnitir
Tar Sands
Tar Sands
 
Posts: 894
Joined: Sat 02 Apr 2005, 04:00:00
Location: Down Under
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby UFCjunkie » Thu 31 May 2007, 08:11:07

To me this sounds like the war between the humans and the machines.

I believe the machines would wipe the waste (us) out when they don't need us anymore, why wouldn't they? Because they need pets?, food?

I believe we will be usless and be killed one way or another.

I think that heading to singularity is giving up the fight without even fighting. I think this society realy need a time of anarchy where we humans destroy everything that can achieve singularity. Because to me it sounds like the end of humans and the begining of the machines or whatever. I'm a human I want to be human I will fight for the humans. I want anarchy until this mad scientists are dead and gone. Who need that technology? Not the humans, maybe the coming race but I'm a human I want to live like a human.

I hope that the economic crash and peak oil will put a stop to this progress and that billions of hungry people will see there power over small govenrments. The armys are the people they wont kill there families, because they would have to.

This is my dream though, that man will once again take control. Fuck the machines, they are for the weak. Fight the machines because they will fight us.


$this->bbcode_second_pass_quote('', 'L')et's just hope that the military doesn't build Skynet before a university or business builds Friendlynet. Shocked

We can do something about it now or we can hope and then get killed later. What would you think? The military or friendly, who will be first? even if the friendly is first the military would take over.

This is the most dangerous road we ever walked on.
I Love This Planet, I Don't Want To Kill It Anymore!

Man Belongs To The Earth * Earth Do Not Belong to Man

27/3-07 The Day UFC Won The War!
User avatar
UFCjunkie
Peat
Peat
 
Posts: 70
Joined: Thu 08 Feb 2007, 04:00:00
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Omnitir » Thu 31 May 2007, 19:51:36

"The human mind likes a strange idea as little as the body likes a strange protein and resists it with a similar energy."
- W. I. Beveridge








$this->bbcode_second_pass_quote('Shannymara', '')$this->bbcode_second_pass_quote('Omnitir', 'I') humbly bow before your obvious wisdom that can only exist with such a large post count.

Have some respect for your elders. Jeez. What's this world coming to?

Why should I show respect to someone that shows me none and treats everything I say, no matter what I say, with outright contempt? An expert on agriculture who couldn't contribute to a discussion about GM crops anything more that "these guys are jerking each other off"?

I won't show respect to such a person simply because they have been around for a decade or two longer than me.
"Mother Nature is a psychopathic bitch, and she is out to get you. You have to adapt, change or die." - Tihamer Toth-Fejel, nanotech researcher/engineer.
User avatar
Omnitir
Tar Sands
Tar Sands
 
Posts: 894
Joined: Sat 02 Apr 2005, 04:00:00
Location: Down Under
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby rsch20 » Fri 01 Jun 2007, 00:49:33

$this->bbcode_second_pass_quote('TheDude', 'O')ne would certainly hope they're not working on a superpowerful AI version of Dick Cheney. Which is related to my point regarding the Haves sequestering themselves from us Have Nots. Bit of a philosophical minefield here.
I found a post from a few years back on Peak Oil and the Omega Point.


heh it's funny to see you link that, that was my first attempt at bringing this topic up here, it's a good thread imo since I go more into explaining the concept.

$this->bbcode_second_pass_quote('johnmarkos', 'Y')eah, I was reading recently that our brains give us the same biochemical rewards for altruistic behavior that they do for food and sex. If we can reproduce a similar reward system in bots, maybe they won't kill us all. They'll feel good when they help us. So all we have to do is make the AIs altruism fetishists, just like humans, maybe more so. Make them extra good and extra smart.

Of course, once they start building better versions of themselves, they could rip out those silly three laws and come up with their own morality. Nonetheless, if they start out with a desire to be good hard wired in their little positronic brains, maybe we have a chance of that morality influencing the designs they make, too.


This winds up being a technical question more than anything else, even if we plan as much as possible and try to influence a good outcome it probably won't have much effect imo, as you say once it designs new versions it will probably do away with any restrictions.

Thats another reason I find the human-machine merge concept more interesting, for one it leaves us in the picture, but presents a much more complicated picutre, rather than just wondering how the intelligence explosion will work out more social and spiritual factors come into play like reconciling all the cultural divides and opposing viewpoints, and questions of individual free will versus the unified mind (yes i'm talking borg, but please rid yourself of that pre-conception and consider what a 'internet' of minds would really be like).

In many ways it's much more unworkable to think about than AI, and invites more ridicule, but if AI is the route we take I think the risk of becoming 'obsolete' is increased.
User avatar
rsch20
Lignite
Lignite
 
Posts: 275
Joined: Mon 26 Sep 2005, 03:00:00
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby johnmarkos » Fri 01 Jun 2007, 03:32:20

$this->bbcode_second_pass_quote('rsch20', '')$this->bbcode_second_pass_quote('johnmarkos', 'Y')eah, I was reading recently that our brains give us the same biochemical rewards for altruistic behavior that they do for food and sex. If we can reproduce a similar reward system in bots, maybe they won't kill us all. They'll feel good when they help us. So all we have to do is make the AIs altruism fetishists, just like humans, maybe more so. Make them extra good and extra smart.

Of course, once they start building better versions of themselves, they could rip out those silly three laws and come up with their own morality. Nonetheless, if they start out with a desire to be good hard wired in their little positronic brains, maybe we have a chance of that morality influencing the designs they make, too.


This winds up being a technical question more than anything else, even if we plan as much as possible and try to influence a good outcome it probably won't have much effect imo, as you say once it designs new versions it will probably do away with any restrictions.

Maybe, although if we figure out how to engender morality in AIs, they will want to create moral designs as well. I was thinking about this question today while I was biking home from work. I decided that if morality is based on pleasure and pain, moral AIs will have to experience (or at least remember) pleasure and pain as well. If we want them to share our morality, we'll need to make them as human as possible.

Anyway, I was thinking today that extension of human abilities does give this project a head start; it also ensures a basis in human experience. Nevertheless, for the time being, independent AI has the advantage that it can evolve much faster than we can.

I was also thinking about what the experience of consciousness is like in terms of a computer. Is consciousness an application or a process? It's not a process like any I've ever seen, because it runs continually for decades. I decided that there really isn't any appropriate software metaphor. Some people have suggested that because there's no one thing there -- a person is a collection of a huge number of cells -- that consciousness is an illusion. I disagree with that assertion.

The metaphor for selfhood that I like is that of an army, which certainly exists and acts but is made up of thousands of independent entities.
User avatar
johnmarkos
Tar Sands
Tar Sands
 
Posts: 866
Joined: Wed 19 May 2004, 03:00:00
Location: San Francisco, California
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby johnmarkos » Fri 01 Jun 2007, 04:04:16

$this->bbcode_second_pass_quote('Omnitir', '')$this->bbcode_second_pass_quote('johnmarkos', '')$this->bbcode_second_pass_quote('Omnitir', 'I')t won't necessarily think like we do. It depends how we build it. If it's built the right way, it could naturally desire to help people.

Yeah, I was reading recently that our brains give us the same biochemical rewards for altruistic behavior that they do for food and sex. If we can reproduce a similar reward system in bots, maybe they won't kill us all. They'll feel good when they help us. So all we have to do is make the AIs altruism fetishists, just like humans, maybe more so. Make them extra good and extra smart.

Of course, once they start building better versions of themselves, they could rip out those silly three laws and come up with their own morality. Nonetheless, if they start out with a desire to be good hard wired in their little positronic brains, maybe we have a chance of that morality influencing the designs they make, too.


I agree with your first paragraph but have doubts about the second. You're right about how we need to design them to be inherently good from the beginning. But I think the Asimov "three laws safe" concept has distorted common concepts of AI somewhat. As you say, the three laws are silly. Asimov made them up for dramatic purposes and showed how complex it is to make an AI safe - the three laws don't work. But the notion that an AI will re-write it's morals I think is also misguided. As it rewrites it's code to become more intelligent, meaning to calculate quicker, why would it change it's basic purpose?

I think if an AI was designed from the beginning to be a certain way, then that is what it is, and no matter how advanced it becomes, as it rewrite it's code to increase it's performance, it will not spontaneously change it's basic premise from which it was originally designed.


Have you read Cory Doctorow's story, I, Robot? He has done a series of stories with the names of classic science fiction works. In this story, North American robots are inferior designs, monopoly-built and controlled by Asimov's three laws. Eurasian models, on the other hand, are independent beings with free will and their own morality. Though they are powerful, they wouldn't rewrite their own morality any more than you or I would ours.

My point is that AI would need human-like morality, and to get that, its experience of the world would need to be as human-like as possible. I think roboticists should start working on taste buds so that AIs can experience gustatory joy. Maybe we're already moving in that direction.
User avatar
johnmarkos
Tar Sands
Tar Sands
 
Posts: 866
Joined: Wed 19 May 2004, 03:00:00
Location: San Francisco, California
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Plantagenet » Fri 01 Jun 2007, 17:13:51

$this->bbcode_second_pass_quote('johnmarkos', '
')My point is that AI would need human-like morality, and to get that, its experience of the world would need to be as human-like as possible


AI doesn't "need" anything. Robots are not alive and not human.

If we want them to have human-like morality or if we want them to be constrained by Asmovian rules, then we need to design it in and hard-wire it in ourselves. There is no way to devise an "education" or series or experiences that would guarantee a humane, nice, friendly, AI anymore then there is way to educate children to guarantee a humane, nice, friendly adult human.
Never underestimate the ability of Joe Biden to f#@% things up---Barack Obama
-----------------------------------------------------------
Keep running between the raindrops.
User avatar
Plantagenet
Expert
Expert
 
Posts: 26765
Joined: Mon 09 Apr 2007, 03:00:00
Location: Alaska (its much bigger than Texas).
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby johnmarkos » Fri 01 Jun 2007, 19:36:32

$this->bbcode_second_pass_quote('Plantagenet', '')$this->bbcode_second_pass_quote('johnmarkos', '
')My point is that AI would need human-like morality, and to get that, its experience of the world would need to be as human-like as possible


AI doesn't "need" anything. Robots are not alive and not human.


It needs morality so that it doesn't kill us.

$this->bbcode_second_pass_quote('', 'I')f we want them to have human-like morality or if we want them to be constrained by Asmovian rules, then we need to design it in and hard-wire it in ourselves. There is no way to devise an "education" or series or experiences that would guarantee a humane, nice, friendly, AI anymore then there is way to educate children to guarantee a humane, nice, friendly adult human.


I agree on the guarantee part. Some people are immoral, or amoral. Nevertheless, those people are either sociopaths or they are violating an internalized human morality (which, I grant, differs somewhat from person to person). Reacting to recent research that shows that being "good" causes a pleasurable response in the brain like that experienced during food or sex, I posited that human morality is based on pleasure and pain. That is, morality is an evolutionary adaptation that favors the survival of the species. We learn morality by figuring out which behaviors make us feel good. I posit that to be moral, an AI would need to experience pleasure and pain.

Anyway, Omnitir and I agreed to toss the three laws out the window because they're too simplistic for a hypothetical sophisticated AI. Such a being could just reprogram itself around the three laws if it wanted too. However, an AI with a deeply ingrained morality like or better than that of humans would not reprogram itself to be immoral. It would desire to be good. In addition, it would want to design AIs that were equally moral or better.

The Age of Virtuous Machines, by J. Storrs Hall
User avatar
johnmarkos
Tar Sands
Tar Sands
 
Posts: 866
Joined: Wed 19 May 2004, 03:00:00
Location: San Francisco, California
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby TheDude » Fri 01 Jun 2007, 21:34:13

$this->bbcode_second_pass_quote('johnmarkos', 'I') posit that to be moral, an AI would need to experience pleasure and pain.


The first analogy that came to mind is a cop being ordered to taser themselves, before they'll allowed to carry one. Doesn't convey what you're talking about though, I don't think. Didn't someone mention how a human has go through a multi-year growing process? I'd hate to be around a strong AI in the Terrible Twos!
Indeed I'm free to whatever I want right now - clean house, press the Submit button (DON'T DO IT!), anything. It's strange to imagine an AI interacting with [s]the rest of [/s]humanity (little anthropomorphization slip there), trying to solve its problems. I know it would immediately have all the answers, how would it impose them?
Baffling stuff. I came across an interesting term today, "steady-state economy." Yep, Einstein's biggest mistake ever applied to the world. Well, sort of. Actually, not so, according to its proponents. Permaculture writ large. That's the Utopia you hear a lot about here.
Cogito, ergo non satis bibivi
And let me tell you something: I dig your work.
User avatar
TheDude
Expert
Expert
 
Posts: 4896
Joined: Thu 06 Apr 2006, 03:00:00
Location: 3 miles NW of Champoeg, Republic of Cascadia
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Plantagenet » Fri 01 Jun 2007, 21:55:09

$this->bbcode_second_pass_quote('johnmarkos', 'R')eacting to recent research that shows that being "good" causes a pleasurable response in the brain like that experienced during food or sex, I posited that human morality is based on pleasure and pain. That is, morality is an evolutionary adaptation that favors the survival of the species. We learn morality by figuring out which behaviors make us feel good. I posit that to be moral, an AI would need to experience pleasure and pain.



So are you proposing to design the AIs brain so it feels pleasure when it does things you think are good, and it will feel pain and stop doing the things you think are bad?

Ultimately, thats no different then the Asimovian approach, except he would program the AI so it was incapable of doing certain bad actions and you would program the AI so it would feel pain and stop if did certain bad actions. [smilie=4robot.gif]
Never underestimate the ability of Joe Biden to f#@% things up---Barack Obama
-----------------------------------------------------------
Keep running between the raindrops.
User avatar
Plantagenet
Expert
Expert
 
Posts: 26765
Joined: Mon 09 Apr 2007, 03:00:00
Location: Alaska (its much bigger than Texas).
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby johnmarkos » Sat 02 Jun 2007, 00:02:27

$this->bbcode_second_pass_quote('Plantagenet', '')$this->bbcode_second_pass_quote('johnmarkos', 'R')eacting to recent research that shows that being "good" causes a pleasurable response in the brain like that experienced during food or sex, I posited that human morality is based on pleasure and pain. That is, morality is an evolutionary adaptation that favors the survival of the species. We learn morality by figuring out which behaviors make us feel good. I posit that to be moral, an AI would need to experience pleasure and pain.



So are you proposing to design the AIs brain so it feels pleasure when it does things you think are good, and it will feel pain and stop doing the things you think are bad?

Ultimately, thats no different then the Asimovian approach, except he would program the AI so it was incapable of doing certain bad actions and you would program the AI so it would feel pain and stop if did certain bad actions. [smilie=4robot.gif]


Getting warmer.

I would program the AI so it would feel guilt and anguish when it did bad actions and have pleasant thoughts when it did good ones -- same as humans. I think morality is *based* on pain and pleasure but it doesn't literally cause physical pain or pleasure.

The point is to engender a morality that the AI actually believes in. The problem with the Asimovian approach is that a sufficiently clever robot can get around the three laws. A moral robot wouldn't try to reprogram its own morality, at least not to become more evil.
User avatar
johnmarkos
Tar Sands
Tar Sands
 
Posts: 866
Joined: Wed 19 May 2004, 03:00:00
Location: San Francisco, California
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby Plantagenet » Sat 02 Jun 2007, 01:03:29

$this->bbcode_second_pass_quote('johnmarkos', ' ')A moral robot wouldn't try to reprogram its own morality, at least not to become more evil.


Why not? The robot may think its moral to do things you think are evil.

Your premise assumes (1) that there is an absolute morality that intelligent robots and people would agree upon and (2) that robots would perceive the exact same things and actions to be moral or evil that you do.

In actuality, morality is relative. For instance, premarital sex was considered to be highly IMMORAL in 18th century England and completely MORAL in 18th century Tahiti. Genital mutilation is considered EVIL today in the west, but is considered a sign of great MORALITY in parts of Africa today. Suicide and mass murder are considered EVIL in most of the world, but are so highly laudable and praiseworthy in some Islamicist terror cults that the suicide-murderer is considered a holy martyr who will be rewarded for all eternity with 72 puncture-repairing virgins.

There is no absoluate morality, and it is naive to assume an AI robot who was free to make up his own "mind" would find the same things to be immoral and moral that you do. 8)
Never underestimate the ability of Joe Biden to f#@% things up---Barack Obama
-----------------------------------------------------------
Keep running between the raindrops.
User avatar
Plantagenet
Expert
Expert
 
Posts: 26765
Joined: Mon 09 Apr 2007, 03:00:00
Location: Alaska (its much bigger than Texas).
Top

Re: Techno-Future ramblings (psst Omnitir)

Unread postby UFCjunkie » Sun 03 Jun 2007, 16:45:53

These last few days I've been thinking about this new, modern, high tech world some people here talks about.

I just can't get it that it will be good for us. We can't even possible know how it's going to be like. I think it will be a even more mad world than it is now. But we can't imagine how mad until we're there.

I think we will find out that we have lived in an illusion that we are goint to save the world with tech but in the real world we have built our killers. We will by ourselfs build the next race that will wipe us out. Nature couldn't make a animal that would wipe us out we would kill it before it reach any stage where it began to be a treath. Instead we will build it our self in belief that we build a savior to the world, but infact we build our own executioner. We wont realise it until it's too late and when it's a fact.

I can't imagine another outcome. And the road to our extermination will only get nastier and nastier. This will be our fuel to keep on dreaming about this high tech that is going to "save" us.

We humans will more and more leave our humanity and we will more and more loose our human feelings and meaning. We wont any longer have our male and femal standards, male and femal will have the same positions and will almost look identical, body and style. We will all be faggots or bisexuall all others will be narrow minded and "old" way of thinking. I also think sience will be mutch involved with births and birthcontrol.

We wont have much privacy and we will be followed every step we take and if the governments can track us when ever they want to, how long back they want. We will all be more in the system than we ever could imagine possible. The crime rate will surely fall to a alltime low thanks to our more controlled society we ever seen. This will be one of the factors that will keep this illusion seem real.

Many people see a high tech future as a better way to live and a way out of our problems. In fact we will be in more problems than ever and we will feel more like prisoners than ever.

But I think we have a choice. We can stop this from happening. We can change the future from being the end of humanity as we know it. We who still want to live like nature has shaped our bodies and minds, can survive and live like we are suppose to. All we have to do is stop the society from go any further in to its self destructed way of life. We have the greatest chance on earth soon with peak oil. Now finally we have the chance to hit back on the society. Thousands of years in captivity will set free when the governments loose their vicious power. If we don't react now we will follow humanity to its grave. We still do have a chance to stop this madness.

These coming years we have to be ready! When the shit goes down we better be ready. Anarchy have to rise and rule the world for some time before all of the society is burnt to the ground in ashes. Then we can rebuild our human race and be stronger than ever before without hightechnology effecting our bodies and lives. If we don't react we will be captives for all of our lives and we will die as captives.

Don't you belive me that this is the way it going to be? No? Prove me wrong or at least give me a reason why the techno society don't will look like I have tried to describe above.

So this is my 2 cents of what is going to happen in the future. But the thing is that I think the ONLY way for us to live is in and with nature. All other kinds of lives is not for us humans and if we choose anothyer way of life, our consequence will be death.

$this->bbcode_second_pass_quote('', 'T')here is no absoluate morality, and it is naive to assume an AI robot who was free to make up his own "mind" would find the same things to be immoral and moral that you do.

I guess you right there...We humans have all different kinds of morality but it don't mather what we think our do. There is only one law that has to be followed and that's the law of the nature, the law of physics. All other is tomato, tometo. We don't need to create a new world for everyone to live in with the "right" standards and morality. It's not our job. If we created robots who can think by them selfs the would think like nature because thats the real way the only working way.

And all "god created us and evolution is stupid" guys...hehe what ever you say.

I say Anarchy! Then let nature set the laws not us god damet, if we do we only set loose hell on earth.

Paradise - Anarchy - Technology society - High technology society - Anarchy - Paradise

Be ready....
I Love This Planet, I Don't Want To Kill It Anymore!

Man Belongs To The Earth * Earth Do Not Belong to Man

27/3-07 The Day UFC Won The War!
User avatar
UFCjunkie
Peat
Peat
 
Posts: 70
Joined: Thu 08 Feb 2007, 04:00:00
Top

PreviousNext

Return to Open Topic Discussion

Who is online

Users browsing this forum: No registered users and 3 guests