Future Tech ... on nanobots and the human soul in the 26th Century

POSTED BY: Karnejj
UPDATED: Thursday, January 5, 2006 21:49
VIEWED: 13373
PAGE 7 of 8

PREV PAGE 1 2 3 4 5 6 7 8 NEXT PAGE

CITIZEN

Wednesday, January 4, 2006 12:19 PM

Quote:
Originally posted by Karnejj:
A large point you seem to be sticking to is that approximations are bad. I'm sure you would possibly attribute "some sense" of consciousness to dogs, maybe even a squirrel. Those brains are definitely not as complex as human brains. They are, in essence, approximations that still give rise to sentience.

Not exactly. Approximations aren't bad per se, but a Dogs Brain is not an approximation or a simulation of the Human Brain. My point is that a simulation is not the real thing; it's a simulation of the real thing, just as my earlier mentioned trip to Jupiter in Celestia isn't a real trip to Jupiter, but a simulation of one.

Quote:
I suppose I can't make you believe in this particular implication of quantum physics.

I've never heard of that as an implication of Quantum physics. The Heisenberg’s Uncertainty Principle seems to refute this by it's very existence, digital systems being deterministic.

Quote:
I see no need to "model subatomic behavior" to model a neuron accurately.
The number of neurotransmitters is guaranteed to be a much smaller value and is more likely the limiting factor in precision, but we'll figure that the number electrons is the necessary limit of precision.

Well you’re moving down to the subatomic by looking at electron count's rather than their cumulative effect. I'd of thought you'd need to model their behaviours to get true accuracy.

Quote:
As I understand it, the first question is whether the current will be generated or not. The neuron holds a threshold which must be exceeded. This threshold is going to be a voltage with less than a 60-bit number of electrons. No approximation needed so far. Next is the total voltage that gets generated - the "arithmetic," as you call it. Again, this is going to be a particular number of electrons which get motivated to pass - a number of finite precision. Finally, some neurotransmitters will be released and, again, this is a finite number.

Now factor in the possibility of tens of thousands of weighted input signals, weighted temporally and by locality, the calculation is fairly complex and I’m not sure how easy it would be to perform on an IC.

Quote:
I'm working on that ... This enlightening and quite lively little debate has motivated me to knock the dust off my research and coding.

Glad to hear it.

Quote:
That was fairly rude, my apologies. But, you disputed me with something with no known consequence. It'd be like saying that the particular speed of light through organic matter is probably the reason that sentience cannot arise in machines. Sure, it's true about the speed of light, but the consequences aren't even close to being known.

Hmm, we both have assumptions underpinning our arguments, I'd say that the assumption that Quantum effects don’t play a role in consciousness is at least as great as the assumption that they do

.

Quote:
Kinda ... learning and high levels of intelligence, so far, are always found paired with sentient beings. Quantum effects can be produced in a lab within a quite non-sentient metal box.

I don’t know computers have demonstrated learning and problem solving Intelligence, yet haven't demonstrated consciousness.

Quote:
Meat made me do it

Uh-uh, it’s always someone or something else’s fault

.

Quote:
Not to get too "Matrix" on you, but ....
As for simulations, what if you were looking at and talking from a distance to EITHER a person OR his reflection (but he could still hear you). What if you couldn't see the mirror? Would you be able to dispute that either one is sentient? With no evidence as to which is sentient and which is the simulation, do you deny consciousness to both?

Some simulations can be good enough to leave you no choice but to accept the results as sentient...

You already know how I feel about human sentience ... What if YOU are the reflection in a mirror that can't be seen ...

Well a mirror doesn’t produ

KARNEJJ

Wednesday, January 4, 2006 4:44 PM

Quote:
Originally posted by citizen:
Not exactly. Approximations aren't bad per se, but a Dogs Brain is not an approximation or a simulation of the Human Brain. My point is that a simulation is not the real thing; it's a simulation of the real thing, just as my earlier mentioned trip to Jupiter in Celestia isn't a real trip to Jupiter, but a simulation of one.

There were a million different clues on your virtual trip .. what happens when you can't tell the difference?

Quote:
Well you’re moving down to the subatomic by looking at electron count's rather than their cumulative effect. I'd of thought you'd need to model their behaviours to get true accuracy.

Well, the issue of "true accuracy" isn't a requirement I believe is necessary. I don't think 60-bit precision is necessary, either. I was just pointing out that exact precision can be calculated on a digital system, theoretically.

Quote:

Quote:
As I understand it, the first question is whether the current will be generated or not. The neuron holds a threshold which must be exceeded. This threshold is going to be a voltage with less than a 60-bit number of electrons. No approximation needed so far. Next is the total voltage that gets generated - the "arithmetic," as you call it. Again, this is going to be a particular number of electrons which get motivated to pass - a number of finite precision. Finally, some neurotransmitters will be released and, again, this is a finite number.

Now factor in the possibility of tens of thousands of weighted input signals, weighted temporally and by locality, the calculation is fairly complex and I’m not sure how easy it would be to perform on an IC.

Ahh ... so you're starting to see the possibility ... eh ... it's a start

Memory space is cheap .. so whether it's 10 or 10,000 signals, they still all only require an address (and maybe a weighting). Simplest way to model the system is to clock your system X times (X = number of connections) to send out your "electronic neurotransmitters," and then allow neurons to check for new inputs on the Xth clock. Update your neuron on that last clock. Check if it received enough "neurotransmitters" to exceed whatever it's firing threshhold function is and then do it all again. (Again, I've neglected the re-wiring procedures that would be needed.)

Quote:

Quote:
Kinda ... learning and high levels of intelligence, so far, are always found paired with sentient beings. Quantum effects can be produced in a lab within a quite non-sentient metal box.

I don’t know computers have demonstrated learning and problem solving Intelligence, yet haven't demonstrated consciousness.

My offer still stands

Show me ANY entity which you consider to have a high level of intelligence and a capacity to learn general information and I can show you a sentient being.

Quote:

Quote:
Not to get too "Matrix" on you, but ....
As for simulations, what if you were looking at and talking from a distance to EITHER a person OR his reflection (but he could still hear you). What if you couldn't see the mirror? Would you be able to dispute that either one is sentient? With no evidence as to which is sentient and which is the simulation, do you deny consciousness to both?

Some simulations can be good enough to leave you no choice but to accept the results as sentient...

You already know how I feel about human sentience ... What if YOU are the reflection in a mirror that can't be seen ...

Well a mirror doesn’t produce a simulation, it just redirects the visible light, and it’s the same as talking face to face. Moreover it’s not the image that is sentient, but the entity.

You can just as easily picture a robot that mirrors a human's movements. The question still stands .. do you deny them BOTH consciousness, because you know that ONE is a simulation? How do you resolve the issue? Can a simulation be good enough, to be considered sentient. It's the same

CITIZEN

Thursday, January 5, 2006 2:02 AM

Quote:
Originally posted by Karnejj:
There were a million different clues on your virtual trip .. what happens when you can't tell the difference?

Most people can't tell the difference between a dream and reality. I say most people as I've experienced lucid dreaming but it's not quite as simple as that.

Point is we have something there where people can't tell the difference.

Quote:
Ahh ... so you're starting to see the possibility ... eh ... it's a start Memory space is cheap .. so whether it's 10 or 10,000 signals, they still all only require an address (and maybe a weighting). Simplest way to model the system is to clock your system X times (X = number of connections) to send out your "electronic neurotransmitters," and then allow neurons to check for new inputs on the Xth clock. Update your neuron on that last clock. Check if it received enough "neurotransmitters" to exceed whatever it's firing threshhold function is and then do it all again. (Again, I've neglected the re-wiring procedures that would be needed.)

It's not quite that simple. One way the inputs are weighted is on the shape of the Dendrites themselves, which means that no two Neurons will react to the same input signals in the same way. I'm also unsure how this system would take into account the temporal weighting I mentioned earlier.

I'm still not convinced that that system could model the vast time dependent interactions of the Neurons though.

But before we get too over excited about the abilities of modern micro electronics I'd like you to consider that the brain is one of the most densely connected network systems in the known universe.
Somewhat un-intuitively, the more intelligent someone is, the less active the brain will be compared to the brain of someone less intelligent performing the same task.

Quote:
Show me ANY entity which you consider to have a high level of intelligence and a capacity to learn general information and I can show you a sentient being.

In so much as I believe true high level intelligence arises from consciousness your right. Would a silicon computer ever show true high level intelligence?

Consider your John Doe example. Is the Robot Doe acting like John Doe because of his/its consciousness, or does it act that way because that's what John would have done?

Quote:
You can just as easily picture a robot that mirrors a human's movements. The question still stands .. do you deny them BOTH consciousness, because you know that ONE is a simulation?

Do we extend the sentient of the Crypt keeper puppeteer from tales from the crypt to the puppet?

Quote:
there are a lot of loopholes in Asimov's laws

The laws are too simple to contain loop holes. A loop hole would be where a Robot could murder a Human being under certain circumstances without breaking any laws. Loop holes come with complexity.

Take a look at Human criminal laws. All legalese ends up doing is add complexity, it doesn't help to refine the laws, it helps to let people with expensive lawyers get away with it

.

Quote:
especially in regards to human vs. human conflicts and the robots' role.

As I understand it there's no loop hole here, the message is simple, in Human vs. Human conflicts, Robots stay the hell out.

Quote:
"The needs of the many outweigh the needs of the few..." What if there happen to be 10 terrorists and 9 hostages?

There would be numerous rules for applying the laws, but at the end of the day if your going to be employing robots in combat situations the last thing you want to do is given them Asimov style behavioural blocks.

Quote:
A man wants to commit suicide by jumping off the adjacent roof .. there's a beanbag gun next to the robot .. the robot wouldn't dare hurt the person, but the inaction allows the person to come to harm ...

Not sure what you mean by this.

More insane ramblings by the people who brought you beeeer milkshakes!
Remember, the ice caps aren't melting, the water

KARNEJJ

Thursday, January 5, 2006 3:09 AM

Quote:
Originally posted by citizen:
Do we extend the sentient of the Crypt keeper puppeteer from tales from the crypt to the puppet?

Don't we have to if we can't see the strings? I mean flip it over, open it up, put it on every scope we've got, and still .. nope .. no strings?

If not, then it does seem that sentience truly IS unattainable artificially (except by accident), because it would seem that a prime requirement would be that "to be sentient, we must NOT know the source of what makes it work [in this case .. the puppeteer]."

... and about ole Robot John Doe ... you never did hazard a guess at when in particular (as his brain was mechanized) he would lose sentience (if ever).

Quote:
Quote:
especially in regards to human vs. human conflicts and the robots' role.

As I understand it there's no loop hole here, the message is simple, in Human vs. Human conflicts, Robots stay the hell out.

Well, that robot company's stock would drop awfully quickly if the supposedly intelligent machines let their owners get smacked around

Dogs have been taken out and shot for less

Quote:
but at the end of the day if your going to be employing robots in combat situations the last thing you want to do is given them Asimov style behavioural blocks.

It's not so much that the blocks are needed because the robots are INTENDED for combat, the blocks are there JUST IN CASE they go off and decide that THEY want combat.

How bad would it suck for the guy who finally invents sentient AI and then he watches his creation go berserk and kill every living human? ... So, yeah ... maybe some Robot Rules are in order

CITIZEN

Thursday, January 5, 2006 3:42 AM

Quote:
Don't we have to if we can't see the strings?

You can't see the strings it's animatronic, and no.

Quote:
Well, that robot company's stock would drop awfully quickly if it let it's owner get smacked around

We're talking too different things here.
How does the robot know whose side to step in on?

The point is if you start putting in provisos like:
You can't hurt Humans, unless they're terrorists.
Then that's a loop hole. What happens if for whatever reason the Robot sees someone as a Terrorist, they have a toy gun or something?

This actually happened with a Soldier shooting a kid in Northern Ireland because all he saw was the end of a realistic toy gun sticking out a window.

If we're going to leave robots open to making these kinds of mistakes because of loop holes introduced by obfustication of the laws then there is absolutely no point in having the laws in the first place.

I assume you've see the Will Smith film version of I, Robot (not much to do with Asimov at the end of the day, but oh well). Heaping more rules (and really even adding rules of interpretation that are anything but literal) would lead to both the actions of Sonny and the central computer (the name of which eludes me at the moment).

I personally think Asimov’s laws of Robotics are about as perfect as they can be, they're like the Ten Commandments, simple direct and too the point, everyone knows where they stand. Though one of your aforementioned Robotics companies would probably want to put a no stealing law in there

.

If we were to ever create an intelligent race, such as Robots, we may want to revise laws such as Asimov’s that we may imbed at a later date, but I think that revision would take the form of removal, rather than anything else.

EDIT:

Quote:
... and about ole Robot John Doe ... you never did hazard a guess at when in particular (as his brain was mechanized) he would lose sentience (if ever).

When they’ve taken away more of his brain’s non-autonomic function than the rest can make up for.

I’d be willing to bet no matter how good Robo Doe mirrors John Doe’s past personality if you’d look into his/it’s eyes, they’d be dead.

It would be like watching a recording of someone who died; do we prescribe the image on the screen with sentient? I’ve never heard it argued that we do.

Quote:
If not, then it does seem that sentience truly IS unattainable artificially (except by accident), because it would seem that a prime requirement would be that "to be sentient, we must NOT know the source of what makes it work [in this case .. the puppeteer]."

Totally possible. Though it wouldn’t be the first time we’ve created a technology and even used it without really knowing how it worked. If we created a biological brain unique to any that has existed, through genetic engineering of Neurons and building them up, and the brain gained sentient we still created it, yet we still don’t know how it works.

More insane ramblings by the people who brought you beeeer milkshakes!
Remember, the ice caps aren't melting, the water is being liberated.

KARNEJJ

Thursday, January 5, 2006 12:13 PM

Quote:

Quote:
Don't we have to if we can't see the strings?

You can't see the strings it's animatronic, and no.

Of course we can see the "strings" if it's animitronic and we open it up. What if you open it up and still can't explain it's intelligent behavior ... *that* is the question at hand.

Quote:
Originally posted by citizen:
I assume you've see the Will Smith film version of I, Robot (not much to do with Asimov at the end of the day, but oh well). Heaping more rules (and really even adding rules of interpretation that are anything but literal) would lead to both the actions of Sonny and the central computer (the name of which eludes me at the moment).

I personally think Asimov’s laws of Robotics are about as perfect as they can be, they're like the Ten Commandments, simple direct and too the point, everyone knows where they stand. Though one of your aforementioned Robotics companies would probably want to put a no stealing law in there .

Actually, following Asimov's laws would probably MANDATE the end of human civilization.

Premise: Human conflict is inevitable.
Premise: Human conflict causes harm to humans.
Premise: Humans would be harmed if they must be physically restrained from entering conflicts.
Premise: Sleep is harmless.

Therefore, if a robot is ordered to not allow humans to come to harm through inaction (and this rule supercedes all others), then the only logical development is probably to anesthetize us all and keep us asleep until we die.

Maybe loophole is the wrong term, but the above scenario is a pretty serious deficiency. I attempt to cover this deficiency in my revised lasws by adding an unconventional meaning of harm in my Zeroeth law (relating to consciousness [as in awake, not referring to sentience]). I think I'll add something to the effect as a new law after the thrid though ...

CITIZEN

Thursday, January 5, 2006 1:27 PM

Quote:
Of course we can see the "strings" if it's animitronic and we open it up. What if you open it up and still can't explain it's intelligent behavior ... *that* is the question at hand.

That's a different question, but if we have a definable originator of the actions it is reasonable to assume that the puppetry is merely beyond our understanding.

You may find this interesting:

Quote:
The Meta-Law
A robot may not act unless its actions are subject to the Laws of Robotics

Law Zero
A robot may not injure humanity, or, through inaction, allow humanity to come to harm

Law One
A robot may not injure a human being, or, through inaction, allow a human being to come to harm, unless this would violate a higher-order Law

Law Two
A robot must obey orders given it by human beings, except where such orders would conflict with a higher-order Law
A robot must obey orders given it by superordinate robots, except where such orders would conflict with a higher-order Law
Law Three
A robot must protect the existence of a superordinate robot as long as such protection does not conflict with a higher-order Law
A robot must protect its own existence as long as such protection does not conflict with a higher-order Law
Law Four
A robot must perform the duties for which it has been programmed, except where that would conflict with a higher-order law

The Procreation Law
A robot may not take any part in the design or manufacture of a robot unless the new robot's actions are subject to the Laws of Robotics

http://www.anu.edu.au/people/Roger.Clarke/SOS/Asimov.html

Understanding that to harm or injure humanity would mean permanently putting people to sleep.

More insane ramblings by the people who brought you beeeer milkshakes!
Remember, the ice caps aren't melting, the water is being liberated.

KARNEJJ

Thursday, January 5, 2006 6:03 PM

Quote:
Originally posted by citizen:
You may find this interesting:
http://www.anu.edu.au/people/Roger.Clarke/SOS/Asimov.html

Understanding that to harm or injure humanity would mean permanently putting people to sleep.

Well, those aren't strictly Asimov's laws ... the one's you quoted include additions from some Roger Clark guy. But, they are surprisingly similar to mine, so he must be a genius ...

Dang, we're good

great minds and all ...

CITIZEN

Thursday, January 5, 2006 6:13 PM

Yeah, I got that , the important one I think is the zero law, and that was added by Asimov himself, after the fact.

More insane ramblings by the people who brought you beeeer milkshakes!
Remember, the ice caps aren't melting, the water is being liberated.

KARNEJJ

Thursday, January 5, 2006 8:32 PM

Quote:
Originally posted by citizen:
Yeah, I got that , the important one I think is the zero law, and that was added by Asimov himself, after the fact.

I think a "procreation" type law actually is a necessity to any list of laws. What good are the laws if they only apply to the first generation of robots? The 2nd generation just gets to be the ones that plug us into the Matrix.

PREV PAGE 1 2 3 4 5 6 7 8 NEXT PAGE