Why are robots always immortal?

[sinanju]

Re: Why are robots always immortal?

 

Actually, what surprises me is that in sci stories that take place in the future that humans aren't immortal.

 

If there are scientist smart enough to create machines with sentient thought then there should be scientist smart enough to stop the aging process. Plus, I can tell you right now there would be a bigger push for that than making your iPod tell you, that you have horrible taste in music.

 

You see that often in written science fiction, where they can take the time to explore more than one maguffin. Larry Niven's "Known Space" novels and stories all assume the existence of Boosterspice, which allows humans to remain young and healthy indefinitely. The Takeshi Kovacs novels (I forget the author) shows us a universe where human identities can be backed up on cortical stacks embedded in the base of the skull. If you get killed, they download you into a new body. (If you're REALLY rich, you also maintain an off-site backup--people who want you really, truly dead will be sure to destroy the stack as well....)

 

In films and television, they a) want to make the future look mostly like the present only with nifty vehicles, weapons and computers, and don't want to overburden the audience with thinking* so it's not likely to happen.

 

A world in which people were immortal (or simply "unaging") would be radically different from the present. Unless that's the _point_ of the movie, you're not going to see it.

 

*It isn't audiences who can't handle complex ideas and plots--it's the morons who rise to the top in the film industry.

[sinanju]

Re: Why are robots always immortal?

 

Most sci-fi writers have a great deal of difficulty dealing with the social consequences of true human immortality. particularly the fact that there is no way to ever make it cheap enough to be available to the entire human race.

 

And you know this how, exactly, DOCTOR Hopcroft?

 

Maybe the immortality treatment will consist of gene therapy applied by means of a modified virus, transmitted person to person. Release it once and it spreads like wildfire (kinda like "Captain Trips" in THE STAND) until every human on the planet catches eternal youth.

 

That's exactly as likely as your claim that it could never be available to everyone.

 

The assumes it's even theoretically possible, which is doubtful. There may be a certain point in human aging where Alzhiemer's Disease or the equivalent becomes utterly inevitable. We were not hard-wired to live forever -- it makes no evolutionary sense. On the contrary, immortality = little or no reason to have children = complete genetic stagnation.

 

Humans as they exist today may not be indefinitely upgradeable. But we're living at the beginning of human history, not the end. Humanity 2.0 (or 3.0 or 4.0) may well be redesigned to be capable of existing indefinitely.

 

The evolutionary argument is irrelevant. Evolution is blind and plays only at the species level. Once you introduce intelligence, the individuals within the species can choose to begin directing their own advancement. And probably will. Evolution didn't design humans capable of living in the desert or in the arctic, or for months at a time under the sea or in outer space. But we do--because we used our intelligence to make it possible. Why should we stop there?

[Curufea]

Re: Why are robots always immortal?

 

*It isn't audiences who can't handle complex ideas and plots--it's the morons who rise to the top in the film industry.

And television network programming.

[Nyrath]

Re: Why are robots always immortal?

 

And you know this how, exactly, DOCTOR Hopcroft?

 

Maybe the immortality treatment will consist of gene therapy applied by means of a modified virus, transmitted person to person. Release it once and it spreads like wildfire (kinda like "Captain Trips" in THE STAND) until every human on the planet catches eternal youth.

 

That's exactly as likely as your claim that it could never be available to everyone.

Well, I tend to agree with Mr. Hopcroft, but not from a medical standpoint. Instead I agree from a cynical standpoint.

 

Developing such an immortality cure will be very very expensive. The powers that be who could fund such a thing will want a sizable return on their investment. You do not get a sizable return on one's investment if you give the cure away for free. Indeed, the return increases if you intentionally restrict the supply of the cure.

 

Some sort of self-propagating cure is never going to be released into the wild. It is almost impossible to insure that the cure will not have some catastrophically bad result on some segment of the population. Consequences can range from massive class-action lawsuits up to war-crime convictions for biological warfare or genocide. Just imagine if one of the side effects was, say, inducing terminal cancer in everybody under the age of ten.

 

An immortality cure will have drastic effects upon the social order. The political powers that be have a vested interest in maintaining the status quo. It would be in their interest to restrict access to such a cure. Such a cure could ruin such things as the United States system of social security (which depends upon retirees not living very long), retirement planning, and voting demographics.

 

An immortality cure would create an explosive population growth. The only two realistic ways to control this is to either impose mandatory birth control or limit the availbilty of the immortality cure. Which is going to be easier? And no, you cannot ease the population growth by interstellar emigration. Current population increases are about 85 million people a year. This means you have to space lift about two hundred thousand people per day just to stay even.

[Dauntless]

Re: Why are robots always immortal?

 

Food for thought:

 

1) The mind is not the brain

Scientists still mired in deterministic classical physics hold on to the notion that matter is made of stuff. This "stuff" exists whether we are aware of it or not. More importantly, our mental states, so say these materialist-objectivist deterministic scientists, are nothing but epiphenomenons of our brain. In other words, our consciousness is nothing but the result of chemical and electrical signals. Our memory nothing more than a neural network pattern.

 

Trouble is, it just aint so. Instead, it's just as likely that the matter might be a creation of the mind. For example, no direct correlate of memory has been found anywhere in the brain. Furthermore, Quantum Mechanics proves that everything is just a probability until it is observed....by a conscious entity (cameras and measuring devices are not enough as Von Neumann proved).

 

2) No (quantum) copying allowed

If the brain works at all on quantum principles, and studies by Penrose, Hammeroff, and others have shown it's quite possible, then we will never be able to perfectly copy the brain even if you do believe in a epiphenomenal mind. Even if we had nanites that could move an atom at a time, that's not good enough. In Quantum Computation, there's an interesting difference between Shannon bits (binary bits) and Qubits (quantum bits). You can't clone a quantum bit....courtesy of Heisenberg's Uncertainty Principle. dohhh, there goes any possibility of "Ghosting" or consciousness downloading even if you don't believe in a Monistic Idealistic interpretation of Quantum Mechanics.

 

3) Permanent DNA

What is DNA? It is a 4-base language that encodes information. DNA is merely the building block for proteins, which in turn make up what we are. Scientists believe that our DNA has a built in "timebomb" that causes us to age. So we could potentially diffuse this. And what about DNA damage such as radiation or errors in transcript RNA copying? Not a problem....save a copy of your DNA. Get some kind of retrovirus or other nanite to "refresh" your DNA back to its original state. Mind you, that's all theory, but I don't see why it won't be possible in the future.

 

4) Artificial biogates

In computer engineering, everything boils down to a logic gate. A NAND or NOR gate is a complete logic gate, meaning that you can build any other gate out of just NAND gates or just NOR gates. Logic gates are the building blocks that you "program" hardware with. You can think of them as analagous to the ATGC pairs of DNA. Just as ATGC pairs are the "logic" (or language) which describes a protein which in turn describes a function, so too are digital logic gates.

 

Well, we've invented artificial biological gates...in essence artificial DNA. What are biogates good for? Creating biological lifeforms or machines that don't run the danger of "polluting" existing lifeforms with the potential for some kind of transgenic crossovers.

[Curufea]

Re: Why are robots always immortal?

 

All good points. But a bit beyond the scope of what can be easily simulated in a roleplaying game.

[Dauntless]

Re: Why are robots always immortal?

 

Well, points 1 and 2 mean that for hard sci-fi games, there should be no Kurzweilesque consciousness or memory downloadings. That means no ghosts, and no immortality via being uploaded into an android body or cybershell.

 

However, point 3 does lead to a kind of immortality....a permanent youth. You can still die of wounds or disease however.

 

If the question is WHY are robots or AI immortal, I believe it all boils down to our own fear of death. People think that if we can achieve mind or memory downloading into robot bodies, we will effectively become immortal. The human ego doesn't like the idea that it will one day no longer exist, and thus mankind has always liked the idea that there is some part of ourselves that is permanent. This part will survive eternally. The idea of an immortal robot (with an immortal AI) is just the new way of expressing this age old human desire for permanence.

 

I think the more interesting question though is why we are afraid of impermanence and change.

[Curufea]

Re: Why are robots always immortal?

 

Possibly it's another throwback from socialisation and our community evolution. Stability = community or some such.

[Inu]

Re: Why are robots always immortal?

 

2) No (quantum) copying allowed

If the brain works at all on quantum principles, and studies by Penrose, Hammeroff, and others have shown it's quite possible, then we will never be able to perfectly copy the brain even if you do believe in a epiphenomenal mind. Even if we had nanites that could move an atom at a time, that's not good enough. In Quantum Computation, there's an interesting difference between Shannon bits (binary bits) and Qubits (quantum bits). You can't clone a quantum bit....courtesy of Heisenberg's Uncertainty Principle. dohhh, there goes any possibility of "Ghosting" or consciousness downloading even if you don't believe in a Monistic Idealistic interpretation of Quantum Mechanics.

There is an exception to Heisenberg, of course: if you can observe without interacting, you indeed can overcome the uncertainty principle. Such would require superpowers or supertech that is far beyond what we can reasonably imagine, but that's not to say it's utterly impossible. Even Hawking leaves open the possibility of this, although he doesn't give it much chance of happening.

 

Though yes, having such a device would likely make the setting a very, very high tech setting indeed.

[Reality]

Re: Why are robots always immortal?

 

Food for thought:

 

1) The mind is not the brain

Scientists still mired in deterministic classical physics hold on to the notion that matter is made of stuff. This "stuff" exists whether we are aware of it or not. More importantly, our mental states, so say these materialist-objectivist deterministic scientists, are nothing but epiphenomenons of our brain. In other words, our consciousness is nothing but the result of chemical and electrical signals. Our memory nothing more than a neural network pattern.

 

Trouble is, it just aint so. Instead, it's just as likely that the matter might be a creation of the mind. For example, no direct correlate of memory has been found anywhere in the brain. Furthermore, Quantum Mechanics proves that everything is just a probability until it is observed....by a conscious entity (cameras and measuring devices are not enough as Von Neumann proved).

 

If you could prove that, you could win the James Randi Million! As it stands, quantum physicists do not as a rule make the mistake of transfering quantum properties (the behaviour of the very small) onto macroscopic objects like brains. At the macroscopic scale, determinism is very much the way to go. That's where all the evidence points. If not, we'd all freeze to death the instant we stopped looking at the sun. Also, saying something hasn't been found is not an argument for whether or not it will be found. Most every biologist whose work on brain function I have read (or video from documentaries that I have seen) confirms that all behaviour correlates to a region of the brain, not just motor function, but all other aspects of personality save memory, which is stored not in a video-recording style all-or-nothing fashion, but as memory segments relevant to the areas in which they are stored (memory of the spatial dimensions of a place stored in one spot, sequence of events at that location in another, etc).

 

As far as robots, one documentary that I watched a few months ago suggested that the only way to get past the problem of increasing complexity of computer thought (that beyond simple tasks digital "intelligence" fails) was to return to an analog means of building robot brains. Analog brains can easily handle increasingly complex tasks. This makes sense, since all organic brains are analog designs. An AI would have a physical brain, with physical parts, and use digital only for storing information. Its personality would come from its physical configuration, much as in humans, not from the data in its banks. If the memory core were taken from a damaged robot brain and inserted into a different robot brain, you would not get the same robot personality, but a different personality with the memories of the old, broken brain.

[Dauntless]

Re: Why are robots always immortal?

 

Actually, the "Schroedinger's Cat" thought experiment does precisely that....transfer quantum properties to a macro object. As does "Wigner's Friend" if i remember correctly. Von Neumann also proved that the quantum entanglement will pass from one measuring device to another...until somehow, a conscious observer comes along and makes an observation.

 

Actually, we can not "prove" determinism either. Determinism requires that the universe A) has material objects which interact locally that these material objects exist independent of our observation of them and C) that the state data (location, momentum etc) is exact and precise. Quantum theory has shot down all three assumptions. Reality is nonlocal, objects do not exist until they are observed, and state data is probabilistic not exact.

 

Determinism is simply an ontological assumption we make about the universe. All science rests on certain metaphysical assumptions which are unprovable. It is actually more probable that instead of there being an objective, material, and deterministic world which is dualistic in nature (this scientific paradigm has failed to explain consciousness for 2500 years in the West), it is more likely that reality is subjective, idealistic, probabilistic and monistic in nature. But again, these assumptions are unproveable (though they do remove some icky Quantum paradoxes).

 

As for analog computers, computers don't "think", and unfortunately many computer scientists are stuck on the deterministic, materialistic view of intelligence, and thus believe that consciousness is a byproduct of the vast complexity of a connectionist model (the mind is an emergent feature of a vastly interconnected system of neurons). If you think about it, a transistor is an analog device at the input level...it's just that it can only recognize two states, high or low and thus outputs a discrete value. The problem with an analog gate is noise. Since an analog device has theoretically an infinite number of values, it's true value would be incredibly affected by noise in the system. In digital computers, thanks to Shannon and his insights into Information Theory (and thus Error Correction Codes), we can generally speaking ignore such noise in the system. Noise or "entanglement" is also what's holding us back from developing Quantum computers...making sure the state of the qubit doesn't get "mixed" up with the rest of the quantum nature of the computer itself. Noise essentially acts as a random unpredictable variable we can't account for. ECC and other noise reduction techniques eliminate this for us in digital system since it's relatively easy to correct a bit which can only contain two possible values.

[Dauntless]

Re: Why are robots always immortal?

 

I think the upshot of everything I said above is that if you're going for a hard sci-fi world, not even AI is immortal. Of course, that depends on what you mean by consciousness and intelligence and even immortal.

 

Well, let me step into bizarro world and describe a little about one of the lesser known Quantum Interpretations: Monistic Idealism. MI says that there is only one consciousness that pops all qwf's (quantum wave functions). We think we have unique minds and thus consciousnesses, but according to MI, we don't. This solves Wigner's Friend dilemma. There is only one consciousness, and our brains are sort of like "mental receivers" that tune in to it. Moreover, we're stuck at one "frequency". This is why MI adherents say, the brain seems to control the mind. If you break a radio, would you say that the music came from the radio? No, but that's precisely what the materialists are assuming...the brain is the origin of the mind, but that's like saying music comes from a radio. Fiddle around with the electronics of the radio, and you screw around with its reception....same thing with the brain and the mind.

 

I've thought of merging Buddhist and Advaitic philosophies (as well as my own knowledge of Computer Science and AI) into a game world that discusses some of these issues. I guess that's why I liked Battlestar Galactica so much, because to me, maybe AI might be able to see this universal consciousness (God) better than we humans can. Or maybe they will be just as deluded and mired in maya (the hindu word for illusion) by their skhandas (the buddhist word for the 6 senses which reveal only a false and illusory world...the 6th sense by the way is our mind, and thus our mind is not exactly the same as our consciousness, or vijnana, in Buddhist philosophy).

 

And even if a robot isn't truly immortal, it definitely would have the potential to last a lot longer than we V.1 humans can.

[Curufea]

Re: Why are robots always immortal?

 

Loss of identity would be the primary fear of an AI. Corruption of their inner beings - what makes them "them".

[Nyrath]

Re: Why are robots always immortal?

 

I think the upshot of everything I said above is that if you're going for a hard sci-fi world' date=' not even AI is immortal. Of course, that depends on what you mean by consciousness and intelligence and even immortal.[/quote']

And just to add some chaos to this confusion, if someone or something is said to be "immortal", that claim very difficult to prove, but easy to disproved. To prove it you have to wait until the end of time and see if they are still alive.

 

Does one mean "immoral" as "immune to the the ravages of old age forever", or "has an unusually long life-span"? If the latter, are they immortal on the scale of hundreds of years, on the scale of thousands of years, immortal on the scale of millions of years....

[austenandrews]

Re: Why are robots always immortal?

 

I think the upshot of everything I said above is that if you're going for a hard sci-fi world' date=' not even AI is immortal. Of course, that depends on what you mean by consciousness and intelligence and even immortal.[/quote']

We can talk about tasting our own tongues all day, but in the end it boils down to behavior. If AI #1 behaves exactly like AI #2, at that moment they are the same. So if I can make AI #2 act exactly like AI #1, based on the same memories and so on, then AI #2 can be considered an extension of AI #1. Of all the myriad measures of identity, continuity of behavior must stand as the overriding one. Because if you break it all the way down to quantum states, technically you're not the same intelligence in the morning as you were the previous night.

[Dauntless]

Re: Why are robots always immortal?

 

In a sense, you've hit the nail on the head.

 

What are we exactly? Our bodies? Our memories? Our thoughts? Our feelings? This is where the roleplaying aspect comes into play. What are we exactly? I think this is why Transhuman genres have a small fan following (small because it's very deep and philosophical...something many gamers have a problem grasping enough to wrap a roleplaying handle around).

 

If we are solely our memories, then am I merely a storehouse of information? My experiences? In a way, you are right, we are different everytime we wake up. In fact, we are different from moment to moment. Moreover, memories are only half the battle, for all a memory is is stored knowledge. How one acts on that stored knowledge (intelligence and emotions) is something different altogether. As I'm trying to point out, our mind, consciousness, emotions and other mental states are (probably) not purely deterministic classical states described by atoms (they are partially determined by it, but not solely). Thus I am not purely my memories, though my memories may help shape how I evaluate and judge things.

 

Thanks to chaos theory, even the tiniest differences can lead to huge changes. Since you can not perfectly copy something, they will be different.

 

But, I can already hear the argument, we can make exact digital copies. As I mentioned earlier, there's a difference between a Shannon bit and a Qubit. Classic computers work with discrete Shannon bits. I'm not sold on the idea that computers lack true creative intelligence and consciousness simply because they aren't complex enough. The majority of Computer Scientists (who believe Strong AI is possible in the first place) think that if we can just create enough neural networks, a gestalt connectionist intelligence (consciousness) will emerge using our current classical Shannon bit based computers.

 

However, such a system would by necessity be a purely deterministic one. Richard Feynman proved that classical computers can do anything a quantum computer can (albeit eponentially slower) except one....simulate non-locality. If our own brain operates on non-local (and thus quantum) principles, then classical computers will never be able to have consciousness (at least human consciousness...it's possible there's other types of consciousness, but since consciousness is a personal and subjective experience, one can define consciousness as having human qualities).

 

Many have heard of Alan Turing's intelligence test (where if you are chatting on a computer, and you can't tell if the thing responding back is a human or not then it's intelligent). However, there's a difference between intelligence and consciousness. Firstly, John Searle came up with a good argument that invalidates Turing's test called the Chinese Room experiment. In essence, Searle says, a computer is just an algorithm machine...it requires no understanding, creativity or leaps of logic.

 

I for one believe that intelligence does have a quantum basis, as do a couple of other researchers (Penrose, IIRC has found structures in neurons called microtubules which could be the basis for quantum mechanics in the brain). We are thus not just our memories, but the thing that acts on those memories. So, if our brain operates on QM principles, we are all unique and thus uncloneable.

 

It wasn't until I started studying Quantum Computation that I realized that physics and information are the same thing. Ask a physicist to look at a tank filled with gas, and he will think pressure, temperature, entropy and thermodynamics. Ask a computer or information scientist what he sees, and he will see a structure of objects that represent information. Reality is quantum in nature and so it's not just physics, or physical objects that it governs, but information as well.

 

Mind you, this is just my own thoughts and theories. There are others who believe in Strong AI from purely classical machines. If you're into such stuff, check out www.kurzweilai.net. Good resource for any Transhumanist or Extropian campaigns someone might want to run.

[austenandrews]

Re: Why are robots always immortal?

 

Since so much of biology is the massed effect of sub-molecular phenomena, I have no problem with the idea that quantum effects play a role in animal mentality. What that effect is, however, is open for debate. If it's just a randomization factor, there's no reason we can't simulate it on a computer. When you multiply quantum randomization over countless iterations, you get a relatively predictable statistical spread. So it may be just a matter of figuring out the equations.

 

I confess I get a tad skeptical when AI arguments are based on determinism or the lack thereof. I know the raw math depends heavily on that condition, but in reality programmers can only dream of a truly deterministic system. I think they're a purely theoretical construct. Meanwhile I have a vague suspicion that quantum effects have become the new "soul," a magical mystery property used to arbitrarily separate humans and nonhumans.

[Captain Obvious]

Re: Why are robots always immortal?

 

We can talk about tasting our own tongues all day...

 

Okay! Mine's tough, but juicy, with a salty flavor and a metallic tang of an aftertaste.

[Dauntless]

Re: Why are robots always immortal?

 

Well, we can't simulate non-locality on a computer, it's just no physically possible. It's the only thing quantum computers can do that classical computers can not. The cool thing about non-locality is that when you measure one point of a set, you in effect measure all possible permutations of that set (the power set). If our brains have a non-local component to them, then I'd say pretty much all bets of classical AI are off.

 

Quantum Mechanics is so bizarre because it describes everything (except gravity) but we have no idea HOW it works. It just does. You've got your copenhagen intrepetation, your many worlds advocates, the holographic universe proponents, the "universe is a quantum computer" theory, and the one I lean towards, monistic idealism.

[input.jack]

Re: Why are robots always immortal?

 

I like this Dauntless guy. But this conversation got a whole lot more complicated about the time he showed up.....

[Zeropoint]

Re: Why are robots always immortal?

 

My take on the "Chinese Room" scenario is that if the room/operator combination can, in fact, carry on a conversation indistinguishable from that of a human Chinese-speaker, then it follows that the collection of data that the operator is working with DOES constitute an artificial intelligence--one that's running very slowly on grossly inefficient hardware, of course.

 

This would be exactly anaologous to the information making up the data and executable of a traditional computer-based AI.

 

It's definitely counter-intuitive, and maybe even offensive in some way, to say that the collection of books "understands" anything, but if it didn't, would it be able to produce the right output?

 

Two parting thoughts:

 

First, that the Turing Test, as I see it, is based on the realization that no one can ever KNOW what's going on in anyone else's head, and we therefore have no choice but to judge someone's intelligence, and indeed their personhood, by their output. In the case of a human being, output consists of words, tone of voice, facial expression, body language, even odors. However, we recognize other humans as beings like ourselves, and don't bother to evaluate them on this basis: we assume that if WE are a person, THEY must be, too, because they're like us. When dealing with another form of intelligence, we don't have this species chauvanism at work, and are much more critical. The purpose of the Turing Test is then to eliminate any factors by which our inherent meatist prejudices might be invoked, and to cause us to evaluate the entity on the other end solely on the basis of intellectual output.

 

Second, that questions like "what constitutes a person?" and "what is intelligence/consciousness?" are important in their own right, but we are entering a period of history in which the answers to those questions will have practical value.

 

Zeropoint

[JDHMorgan2003]

Re: Why are robots always immortal?

 

I was wondering what one would consider an "aging" attack. I was wondering becuase someone mentioned that robots wouldn't suffer from the effects of aging, and that's why they would buy LS: Immunity to Aging.

 

I've watched several documentaries on airplane crashes where the NTSB concluded that the planes crashes were influenced by metal fatigue. That sounds a lot like an "aging" attack on the metal body of the plane. Another thing that was somewhat important as far as materials science, here in Oklahoma, at least one bridge collapsed due to "aging". The bridge was not maintained properly, and so it got really old, really fast.

[JDHMorgan2003]

Re: Why are robots always immortal?

 

And just my own two cents...I wouldn't allow an immortal robot unless it could function indefinitely without part replacement. Right now, humans can go out and get on a list for part replacement...hearts, livers, kidneys, and other parts are replacable, not easily, but it can be done. Also, humans can get a prosthetic limb, probably not as good as the original, but they're getting better every year. Soon, we might be able to replace via prosthetics that are as good as the original. But we still wouldn't be immortal (LS: Immune to the effects of aging) because the individual components could still wear out.

[Egyptoid]

Re: Why are robots always immortal?

 

Bot's could arguably live forever with parts and maintenance.

Advanced tech that allows a Bot PC could work better than ours.

 

Plus, Bots would remember to maintain themselves properly, unlike us chunky humans.

[JDHMorgan2003]

Re: Why are robots always immortal?

 

But that's just my point. Because the bots have to replace their worn-out (i.e. old, or aged) parts, they are not immortal. So, if someone used a sfx on them that said the parts aged, got old, fatigued, stressed...say as a Drain BODY NND (LS: Immortal)...they would not be protected.