Acceleration Of Technological Innovation Means An Endpoint For Us Within Decades?

[Dave1001]

Kurzweil points out the logical nearest forseeable future point of technological progress is all matter being turned into maximally intelligent matter expanding and converting all other matter at speed asymptotically approaching the speed of light. (What would be unforseeable is whether by then greater insights into the universe will allow the smart matter to expand quicker than the speed of light.)

This point has converted me a bit from being a moderate optimist into a moderate pessimist about our best odds of persistence as conscious, self-reflective entities. At some point, perhaps predictable on Kurzweil's model timeline of technological acceleration, smart matter is going to be able to make more efficient use of 4 pounds of matter than the smartest human can with our brains. This by itself won't necessarily result in the cessation of conscious existence, but it could result in human brains becoming the most unintelligent/unproductive form of matter within this expanding sphere of smart matter. More intelligent matter will have both the incentive to convert human brains into smarter matter, and the cognitive advantage to do so unilaterally.

This by itself won't necessarily result in the cessation of our conscious, self-reflective existence, but it would likely force us from the medium in which we're most sure it does exist.

At a future point, or perhaps at pretty much the same point, our conscious self reflective selves may just be a floating less efficient algorithm, slowing down the expanding sphere of smart matter. Once again it would have the incentive and cognitive advantage to unilaterally repurpose that use of matter energy to its own growth.

I don't see a way out of this problem. Thus, although technological progress is likely to result in solving the traditional causes of mortality in a future which we may live to see, I think it is also likely to result in the cessation of our existence in any sort of conscious, self-reflective way only a few decades later.

To say we continue to exist at that point, would be like saying we existed 10 million years ago. Sure the particles that make us up existed then and will exist in a future without us as conscious self-reflective identities, but functionally, I think we'd no longer exist in any appreciable or experiential way.

I take refuge in the fact that my brain may be dumb enough matter to not see a way out of this expiration point for me or you as conscious, self-reflective entities. ;-)

[TobiasTheViking]

imo that sounds like a lot of goobledegoo.

[Dave1001]

imo that sounds like a lot of goobledegoo.

Thanks. That's reassuring.:)

[DanishDynamite]

Kurzweil points out the logical nearest forseeable future point of technological progress is all matter being turned into maximally intelligent matter expanding and converting all other matter at speed asymptotically approaching the speed of light. (What would be unforseeable is whether by then greater insights into the universe will allow the smart matter to expand quicker than the speed of light.)
No idea who Kurzweil is. As far as his logical prediction goes, it's way out there. At least in the way you describe it.
This point has converted me a bit from being a moderate optimist into a moderate pessimist about our best odds of persistence as conscious, self-reflective entities. At some point, perhaps predictable on Kurzweil's model timeline of technological acceleration, smart matter is going to be able to make more efficient use of 4 pounds of matter than the smartest human can with our brains. This by itself won't necessarily result in the cessation of conscious existence, but it could result in human brains becoming the most unintelligent/unproductive form of matter within this expanding sphere of smart matter. More intelligent matter will have both the incentive to convert human brains into smarter matter, and the cognitive advantage to do so unilaterally.
This is gobblydigook. Unless you can explain this new paradigm better, it will remain so.

[TobiasTheViking]

Thanks. That's reassuring.:)

Sorry, wasn't the intend. What i meant to say was.


OOOOOOOOOOOOOOHHHHHHHHHHHH NOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OMG OMG OMG OMG, IT'S THE END OF THE WOOOO-HUUUURRR--OOOOLLDD.


Honestly, Dave, Look back at the last 50 years. How many doomsdays have we faced from fanatics, the media, and pseudo-scientists?

While some may be the legit(i'm not going to try and wheed through them) most have been total bull.

We had the "we are going into an iceage", which was out as late as the early 90's.
We had nuclear power would kill all life on earth(from tjernobyl, and no, i'm not gonna try with the spelling).
We had the rock flying through space.
We had the other rock flying through space.
We had the pop science magazine talking about the impending doom from a big burst of energy from a supernova.
We had SARS
We had H5N1
We have global warming.

And that is just from the top of my head.


There have been so many, just in my 23 years of life i've heard way too many, and see them fall flat on their faces.. Sorry, i can't take any seriously anymore. Especially nothing as far fetched as the idea in your OP. For now, the idea in the OP is just a philosophical idea. Nothing more, nothing less.

If you wanna fear the OP, there are far more gloom things in store for the human kind. Which are totally unavoidable.

Personally i've stopped following the news, because all it does is try to scare me. It doesn't work, i just get pissed at their attempts. If there is some REAL danger, i'm sure i will find out in due time.

The end of the world is coming, you can be sure of that Dave, but it is a few billions^billions^billions year away. Or possibly more.

[Dave1001]

This point has converted me a bit from being a moderate optimist into a moderate pessimist about our best odds of persistence as conscious, self-reflective entities. At some point, perhaps predictable on Kurzweil's model timeline of technological acceleration, smart matter is going to be able to make more efficient use of 4 pounds of matter than the smartest human can with our brains. This by itself won't necessarily result in the cessation of conscious existence, but it could result in human brains becoming the most unintelligent/unproductive form of matter within this expanding sphere of smart matter. More intelligent matter will have both the incentive to convert human brains into smarter matter, and the cognitive advantage to do so unilaterally.

This is gobblydigook. Unless you can explain this new paradigm better, it will remain so.

Okay. In short. What were our human brains before they were brains? They were elements and minerals in the Earth. With 6 billion brains, 4 pounds each, that's about 24 billion pounds of matter that's more intelligent than it would be otherwise. By intelligent I mean able to reorganize the environment for its persistance. One could look at the evolution of life on the Earth, starting with a relatively few single celled organisms, as smarter matter converting less intelligent matter into smart matter.

Well, right now the most intelligent possible formation of 4 pounds of matter is a functioning human brain. But, that's likely to change as innovation continues to occur. Following the predictions of engineers and scientists like Kurzweil, within decades much more intelligent use of 4 pounds of matter will exist than any human brain. And within decades all matter on Earth will be converted into such smart matter. At that point, the intelligent matter would reasonably examine whatever humans with brains remaining, and figure that it could make better use of those 4 pounds by turning it into smarter matter than our brains. And since it will be more intelligent than us, we won't be able to stop it.

Is that any clearer?

[Dave1001]

The end of the world is coming, you can be sure of that Dave, but it is a few billions^billions^billions year away. Or possibly more.

I hope your right. Because all the aforementioned disaster scenarios were ones I was cautiously optimistic we might be able to innovate our ways out of, to the extent that they were real.

What's insidious about this scenario to me is that the very provess of innovation for the purpose of survival may inevitably mean the cessation of our existence in the relatively near future.

[c186282]

Kurzweil has taken a simple idea and extrapolated way outside of the
collected data set. I'm sure the earth and all that live on it will end
someday. However, I do not think it will be replaced by a big ball of
smart goo.

[DanishDynamite]

Okay. In short. What were our human brains before they were brains? They were elements and minerals in the Earth. With 6 billion brains, 4 pounds each, that's about 24 billion pounds of matter that's more intelligent than it would be otherwise. By intelligent I mean able to reorganize the environment for its persistance. One could look at the evolution of life on the Earth, starting with a relatively few single celled organisms, as smarter matter converting less intelligent matter into smart matter.

Well, right now the most intelligent possible formation of 4 pounds of matter is a functioning human brain. But, that's likely to change as innovation continues to occur. Following the predictions of engineers and scientists like Kurzweil, within decades much more intelligent use of 4 pounds of matter will exist than any human brain.
That is certainly a possibility. I.e., that humans will get smarter. Though I doubt it (given your "decades" horizon).
And within decades all matter on Earth will be converted into such smart matter.
Hahahahah!
At that point, the intelligent matter would reasonably examine whatever humans with brains remaining, and figure that it could make better use of those 4 pounds by turning it into smarter matter than our brains. And since it will be more intelligent than us, we won't be able to stop it.

Is that any clearer?
Yep, it's very clear. And without any foundation whatsoever.

Wait, are you a standup-comedian?

[Dave1001]

That is certainly a possibility. I.e., that humans will get smarter. Though I doubt it (given your "decades" horizon).

Hahahahah!

Yep, it's very clear. And without any foundation whatsoever.

Wait, are you a standup-comedian?

I agree, the within decades is counterintuitive. But Kurzweil's writing on the topic have convinced me to be open about that possibility. The cruz of how that could possibly happen so soon from now is that matter organizing itself into smarter matter does so at a rate proportional to how much smart matter there is, and how smart that matter is. Which should mean that once there is matter that exists that hits certain thresholds of being more intelligent and more ubiquitious than humans, it should replicate itself at blinding speed from the perspective of human observers. This, incidentally, is probably how human reproduction and geographic spread looks from the perspective of most other Earth life. Arguably, there is already matter which is more intelligent and more ubiquitous in certain ways than humans: our computers. Obviously it is not yet smarter than us in other ways. But Kurzweil argues that current trends indicate that this gap will be bridged within a couple decades, and at that point I think its speed of replication and innovation is likely to be at least as blindingly fast to us as our rate of innovation and reproduction is to our primate relatives.

I think it's likely this is already happening with the internet.

[TobiasTheViking]

I can see no possible way this could happen within the theory of evolution. So i think we are quite safe from THIS catastrophe... Yet again.

What mechanism could possibly make this happen?

At most humans will get bigger brains, thus taking up more and more matter from the planet to make those brains.

[DanishDynamite]

I agree, the within decades is counterintuitive. But Kurzweil's writing on the topic have convinced me to be open about that possibility. The cruz of how that could possibly happen so soon from now is that matter organizing itself into smarter matter does so at a rate proportional to how much smart matter there is, and how smart that matter is. Which should mean that once there is matter that exists that hits certain thresholds of being more intelligent and more ubiquitious than humans, it should replicate itself at blinding speed from the perspective of human observers.
How?
This, incidentally, is probably how human reproduction and geographic spread looks from the perspective of most other Earth life. Arguably, there is already matter which is more intelligent and more ubiquitous in certain ways than humans: our computers. Obviously it is not yet smarter than us in other ways. But Kurzweil argues that current trends indicate that this gap will be bridged within a couple decades, and at that point I think its speed of replication and innovation is likely to be at least as blindingly fast to us as our rate of innovation and reproduction is to our primate relatives.

I think it's likely this is already happening with the internet.
I can only assume you are joking. Are you aware that what you just wrote above reads as the standard for woo-woo detection?

[Dave1001]

I can see no possible way this could happen within the theory of evolution. So i think we are quite safe from THIS catastrophe... Yet again.

What mechanism could possibly make this happen?

At most humans will get bigger brains, thus taking up more and more matter from the planet to make those brains.

One could say markets, etc. but I think the baseline desire is security in persistence. What drives computers becoming smaller, more powerful, and more ubiquitous? Where is the natural ending point?

[Dave1001]

Are you aware that what you just wrote above reads as the standard for woo-woo detection?

Thanks for sharing your perspective.

[TobiasTheViking]

Well, computers as they are today will never be able to become sentient.. So untill something happens on that front we are safe.

[DanishDynamite]

One could say markets, etc. but I think the baseline desire is security in persistence. What drives computers becoming smaller, more powerful, and more ubiquitous? Where is the natural ending point?
The drive is the "free market". The end point is determined by Physics.

[DanishDynamite]

Thanks for sharing your perspective.
Sorry, Dave, but you are just way, way out there.

Perhaps you would care to explain how this "intelligent matter" gets created outside of the skulls of intelligent creatures?

[Dave1001]

Well, computers as they are today will never be able to become sentient.. So untill something happens on that front we are safe.

I actually don't think sentience outside of humans is a prerequisite for any of this. In fact, my concern is now that sentience will be an inefficiency that this growing intelligence will erase.

This is not so different from Bill Joy's grey goo scenario. The difference is that we may persist until the point when it will be eminently reasonable for the smart matter that surround us to repurpose the matter in which our algorithms exist.

[Dave1001]

The drive is the "free market". The end point is determined by Physics.

Right, but what forces entities to join "free markets"? I think the baseline drive is security in persistance. Otherwise market participants will outgrow and thus eventually have unchallengeable power over market nonparticipants.

[DanishDynamite]

Right, but what forces entities to join "free markets"? I think the baseline drive is security in persistance. Otherwise market participants will outgrow and thus eventually have unchallengeable power over market nonparticipants.
Once again, I have no idea what you are talking about. Sorry.

[Dave1001]

Sorry, Dave, but you are just way, way out there.

Perhaps you would care to explain how this "intelligent matter" gets created outside of the skulls of intelligent creatures?

It might be good to start with reading Kurzweil's and others writings on this topic, as they already covered this ground, pretty convincingly in my opinion.

I'll explain this once, but I don't want to constantly have to redo this in this thread. Basically, some forms of matter smarter than us already exist (computers can do many types of calculations faster and more accurately than we can). Forms of competition and problem solving, most notably markets, are going to drive making computers, etc. increasingly smarter and increasingly cheaper, until eventually one should be able to buy something the size of a cell phone, smarter than the combined intelligence of humanity in 2006, for less than a penny of today's buying power. Based on Kurzweil's predictions that this will happen in under 5 decades.

[Dave1001]

There are people on the boards who are familiar with these concepts. It would be great to hear you weigh in on this. It seems most current respondents are unfamiliar with previous writings on this topic.

[DanishDynamite]

It might be good to start with reading Kurzweil's and others writings on this topic, as they already covered this ground, pretty convincingly in my opinion.
I promise to read them if you can convince me they are worthy.
I'll explain this once, but I don't want to constantly have to redo this in this thread. Basically, some forms of matter smarter than us already exist (computers can do many types of calculations faster and more accurately than we can).
Indeed they can. But they are not smarter than us.
Forms of competition and problem solving, most notably markets, are going to drive making computers, etc. increasingly smarter and increasingly cheaper, until eventually one should be able to buy something the size of a cell phone, smarter than the combined intelligence of humanity in 2006, for less than a penny of today's buying power. Based on Kurzweil's predictions that this will happen in under 5 decades.In which case he is WRONG.

Memory capacitance has little relevance to intelligence. And the pathetic examples of "AI" currently available are just that.

[DanishDynamite]

There are people on the boards who are familiar with these concepts. It would be great to hear you weigh in on this. It seems most current respondents are unfamiliar with previous writings on this topic.
I too would love to hear these "experts" weigh in.

[kedo1981]

Kuzweil is one of my heroes, he’s a brilliant inventor and futurist and I believe he’s correct that in the not to distant future we will have the ability to “upload” or copy our minds into some kind of computer storage and maybe living in some virtual real world.
But the whole smart goo thing smells like woo woo.
The matter in our brains is not smart; our minds are what they are because of the whole package, the body, pattern recognition, and our senses and so on.
How would nano soup sense anything?

[Dave1001]

I promise to read them if you can convince me they are worthy.

Duly noted.:D

[DanishDynamite]

Kuzweil is one of my heroes, he’s a brilliant inventor and futurist and I believe he’s correct that in the not to distant future we will have the ability to “upload” or copy our minds into some kind of computer storage and maybe living in some virtual real world.
This concept is not obviously wrong.
But the whole smart goo thing smells like woo woo.
Glad you agree.
The matter in our brains is not smart; our minds are what they are because of the whole package, the body, pattern recognition, and our senses and so on.
Couldn't agree more.
How would nano soup sense anything?By using the senses impinged on them by the humans who made them. Next question?

[Dave1001]

The matter in our brains is not smart; our minds are what they are because of the whole package, the body, pattern recognition, and our senses and so on.

Isn't that the same as saying that the matter in our brains is smart because of the way that it is organized? If not, where is the distinction, in your view?

[DanishDynamite]

Duly noted.:D
Great! I now expect you to convince me of their worthiness. :)

[DanishDynamite]

Isn't that the same as saying that the matter in our brains is smart because of the way that it is organized? If not, where is the distinction, in your view?
Of course it is "smart" because of the way it is made in our skulls. And?

[Dave1001]

Great!...expect

... ations ;)

[DanishDynamite]

... ations ;)
;)

[Dave1001]

Of course it is "smart" because of the way it is made in our skulls. And?

So on that we agree.

It was Kedo who wrote


The matter in our brains is not smart; our minds are what they are because of the whole package, the body, pattern recognition, and our senses and so on.

[DanishDynamite]

So on that we agree.
Yes. As does every researcher of the brain.
It was Kedo who wrote
Sorry. I stand corrected.

[Dave1001]

Yes. As does every researcher of the brain.

Sorry. I stand corrected.

Well technically I am open to the possibility that elements of our concious existence lie outside of the known material universe. In particular it is hard for me intuitively to resolve the rich multidimensional solipstic experience of reality with its analog as neurotransmitters bouncing around between neurons. However, Kurzweil in his writings has done a good job convincing me that many elements of our cognitive processes can modeled and explained using just the neurons and their systems in our brain. So I'm on the fence, but I still lean towards suspecting that our solipstic experience of the universe occurs in some medium beyond what we materially observe in the brain.

This point is tangential to my concerns in the OP, except to the extent they may mean this threat to our conscious existence may be less than I currently perceive it to be. But few if any neuroscientists I've encountered seem to share my concerns.

[DanishDynamite]

Well technically I am open to the possibility that elements of our concious existence lie outside of the known material universe. In particular it is hard for me intuitively to resolve the rich multidimensional solipstic experience of reality with its analog as neurotransmitters bouncing around between neurons. However, Kurzweil in his writings has done a good job convincing me that many elements of our cognitive processes can modeled and explained using just the neurons and their systems in our brain.
He has?! I'm amazed!

Or rather, I'm not. All the elements of our cognitive processes can be modeled. For years.
So I'm on the fence, but I still lean towards suspecting that our solipstic experience of the universe occurs in some medium beyond what we materially observe in the brain.
I cannot even begin to understand your set of mind. Everything we know is part of our knowledge and part of the Standard Model.
This point is tangential to my concerns in the OP, except to the extent they may mean this threat to our conscious existence may be less than I currently perceive it to be. But few if any neuroscientists I've encountered seem to share my concerns.
Few, if any neuroscientists, would have any inkling of what your ravings might possibly mean.

[Dave1001]

He has?! I'm amazed!

Or rather, I'm not. All the elements of our cognitive processes can be modeled. For years.

I cannot even begin to understand your set of mind. Everything we know is part of our knowledge and part of the Standard Model.

Few, if any neuroscientists, would have any inkling of what your ravings might possibly mean.

Good point. Most, if not all neuroscientists tend towards your more absolutist framing of things, for example "All the elements of our cognitive processes can be modeled. For years". Your general approach represents the best of skepticism and empiricism.

[kedo1981]

The matter in our brains is not smart or intelligent in any way
If you put it in a blender you would have nothing but a high calorie shake

[Dave1001]

Is there a smart people convention going on right now that's keeping many of the regular posters on these topics off the boards?

[DanishDynamite]

Good point. Most, if not all neuroscientists tend towards your more absolutist framing of things, for example "All the elements of our cognitive processes can be modeled. For years". Your general approach represents the best of skepticism and empiricism.
Most, if not all neuroscientist tend toward evidence.

You have none. Are you surprised at your reception?

[DanishDynamite]

Is there a smart people convention going on right now that's keeping many of the regular posters on these topics off the boards?
Not that I know. I'm afraid you'll have to deal with us average dunces.

[Ladewig]

I've read your post a couple of times and I think I am missing something. Please let me ask a few questions.

More intelligent matter will have both the incentive to convert human brains into smarter matter, and the cognitive advantage to do so unilaterally.

1) Is this "more intelligent matter" in the form of a robot of some sort? If the intelligent matter (i.m.) has the power to "unilaterally make use of human brains," then surely it must have the abilities to transport itself and to attack or subdue humans.

2) Why would it have an incentive to convert human brains in some way? We have pretty powerful brains yet we do not capture dolphins and try to convert their limited brains into something as powerful as a human brain.

At some point, perhaps predictable on Kurzweil's model timeline of technological acceleration, smart matter is going to be able to make more efficient use of 4 pounds of matter than the smartest human can with our brains.

I am not sure if you are suggesting that the i.m. will hook us up to a mainframe (ala The Matrix) or if the i.m. is going to kill us in order to make use of the raw materials inside our heads. If the former, then

3) does this i.m. know how sloppy, how inaccurate, and how easily-fatigued the average human's brain is? Considering the difficulty of simply designing a useful interface, wouldn't it be infinitely more easy and efficient for the i.m. to make silicon chips than to hook people to some machine?

if the latter, then

4) wouldn't it be easier to use dog brains or cow brains than people brains?

5) hasn't this concept already been covered in scores of science fiction short stories and television episodes?

[Apathia]

I'm not one of the inteligensia here, but I'll throw in some brass coins.

It's great Science Fiction fodder. Artificially built minds with better, more efficiant hardware, take charge of the environment along with those outdated carbon units (us). Said carbon units are upgraded as much as possible but eventually scraped, while matter at large is used to build better and better hardware components for the second generation sentience.

Organic brains become extinct.

The big questions:

Is Strong A.I. outside an organic environment even possible?

If so will we be able to transfer our minds over into an inorganic brain without losing our humanity?

Will inorganic minds even care?

Are they already out there, heading in our direction?

Has Gregory Benford already made SF with this?

[Apathia]

Yeah, You have to be into Sci-Fi to get this kind of thread. lol

[Dave1001]

I've read your post a couple of times and I think I am missing something. Please let me ask a few questions.



1) Is this "more intelligent matter" in the form of a robot of some sort? If the intelligent matter (i.m.) has the power to "unilaterally make use of human brains," then surely it must have the abilities to transport itself and to attack or subdue humans.


I deliberately left it open, much like Kurzweil does because it could take a variety of forms. The primary principle is that it is matter organized and replicates itself using available less intelligent matter in its surroundings. According to this principle plants are smarter matter than mineral-rich dirt. Humans are smarter matter than plants. At some point if its going to convert our brains into subcomponent material for itself, yes it will have to "attack or subdue humans". I don't picture it being done with whirling buzzsaws, though.:p


2) Why would it have an incentive to convert human brains in some way? We have pretty powerful brains yet we do not capture dolphins and try to convert their limited brains into something as powerful as a human brain.


Well, in a sense we do, although probably in a wasteful and unecessary way. We currently catch dolphins turn them into subcomponents of tuna and eat them. Smarter matter than us wouldn't necessarily have an immediate incentive to convert human brains in some way. But at some point (fairly quickly, according to Kurzweil, because smart matter will start growing phenomenally fast from the 2006 human perspective as it leaves our frame of reference behind on its exponential growth curve) smarter matter than us will have converted EVERYTHING ON EARTH into matter that, like itself, is smarter per mole of particles than a human brain. At that point, it will have the incentive I mentioned to convert our human brains into smarter matter too. And it will be so much smarter than us that we'll likely be powerless to stop it.


I am not sure if you are suggesting that the i.m. will hook us up to a mainframe (ala The Matrix) or if the i.m. is going to kill us in order to make use of the raw materials inside our heads. If the former, then

3) does this i.m. know how sloppy, how inaccurate, and how easily-fatigued the average human's brain is? Considering the difficulty of simply designing a useful interface, wouldn't it be infinitely more easy and efficient for the i.m. to make silicon chips than to hook people to some machine?

if the latter, then

4) wouldn't it be easier to use dog brains or cow brains than people brains?


The latter. Yes it would possibly be easier to use dog brains or cow brains. They'd put up less resistance. But at some point all dog brain and cow brain has been converted to smarter matter than us. At that point the incentive returns. When it came for Fido I said nothing. Then it came for me, and there was no on left to bark.:p


5) hasn't this concept already been covered in scores of science fiction short stories and television episodes?

I'm sure. I'm not making an ego-based play to claim I'm the first person to ever think these thoughts. I'm just saying that reading Kurzweil's piece made this possible future outcome salient to me. Prior to reading his piece, I thought technological innovation could lead to a sort of manmade heaven of youthful physical immortality. Now I think it may be more likely that only a couple decades after that it may lead to a cessation of our existence as sentient beings.

[Dave1001]

I'm not one of the inteligensia here, but I'll throw in some brass coins.

It's great Science Fiction fodder. Artificially built minds with better, more efficiant hardware, take charge of the environment along with those outdated carbon units (us). Said carbon units are upgraded as much as possible but eventually scraped, while matter at large is used to build better and better hardware components for the second generation sentience.

Organic brains become extinct.

The big questions:

Is Strong A.I. outside an organic environment even possible?


Kurzweil makes a strong case that it is. I don't think it's too relevant to my OP though, because smarter matter could be made up of organic subcomponents.


If so will we be able to transfer our minds over into an inorganic brain without losing our humanity?

Will inorganic minds even care?


I'm still skeptical that we can transfer our minds over into an inorganic brain (or a better organic brain) without losing our conscious, experiential existence. It could just be the equivalent of a smarter twin, where we never actually see the world through their eyes, although the world mistakes them for us.


Are they already out there, heading in our direction?


Good question. We seem to have no evidence of that. Kurzweil thinks that means inevitably that we're alone -or rather ahead of the pack technologically- because SETI's done enough work that we'd have detected other civilizations by now. But I don't think he backs that up rigorously. I think he shows that acceleratingly intelligent life/matter is probably rarer than we would think it should be based on the large number of planets with earthlike conditions out there. But rarer doesn't mean non-existent. There could be a sphere of smart matter expanding near the speed of light several thousand light years away from us. We probably won't be able to detect it for awhile.


Has Gregory Benford already made SF with this?

Don't know him by name, but since it's such a specific question I'm guessing the anwer is yes? Tell us more please. :)

[Foolmewunz]

Is there a smart people convention going on right now that's keeping many of the regular posters on these topics off the boards?

What an arrogant crock of doo doo that statement was! :mad:

I'd be happy to discuss Kurzweil or Benford, (whom I believe is responsible for a corollary to AC Clarke's 3rd Law if it's the same Benford), but not with someone so obviously more intelligent than myself and my friends on this thread. I jus' too 'fraid you'd use some of those big confusing words... :D

So, I'll leave you the floor. Maybe Coberst can come over (and bring The Scientologist from a Gravy thread), and you smart guys can all sit around and play mental patty-cake.

[Pidge]

IF YOU SEE ME POSTING HERE, BE MY FRIEND, KEEP ME FROM PROCRASTINATING, AND REMIND ME TO GET BACK TO WORK. THANKS!

Dave1001, back to work!

[marting]

What I want to know is what will happen if "smart matter" gets religion?
Yee haw.

[Mashuna]

What I want to know is what will happen if "smart matter" gets religion?
Yee haw.

Have you been watching Battlestar Galactica again ;)

[Dave1001]

What an arrogant crock of doo doo that statement was! :mad:

I'd be happy to discuss Kurzweil or Benford, (whom I believe is responsible for a corollary to AC Clarke's 3rd Law if it's the same Benford), but not with someone so obviously more intelligent than myself and my friends on this thread. I jus' too 'fraid you'd use some of those big confusing words... :D

So, I'll leave you the floor. Maybe Coberst can come over (and bring The Scientologist from a Gravy thread), and you smart guys can all sit around and play mental patty-cake.

Eh, using big confusing words is not a sign of intelligence. It's probably more of a counterindicator.

:boxedin:

So, what do you think of the ideas in this thread, outside of the arrogant crock(s) of doo doo?

[Soapy Sam]

We talking Vernor Vinge style "Singularity "here?
http://en.wikipedia.org/wiki/Vernor_Vinge

I don't actually see the logic of it myself. "Intelligent matter "is an oxymoron. Intelligence is an aspect of behaviour. A process. All the intelligence in the universe has no mass at all.
If what you mean is that humans may be replaced by their tools, well, possibly. I see that as a reason to make good ones.

[Dave1001]

We talking Vernor Vinge style "Singularity "here?
http://en.wikipedia.org/wiki/Vernor_Vinge

I don't actually see the logic of it myself. "Intelligent matter "is an oxymoron. Intelligence is an aspect of behaviour. A process. All the intelligence in the universe has no mass at all.
If what you mean is that humans may be replaced by their tools, well, possibly. I see that as a reason to make good ones.

No need to get caught up with the words "intelligent matter". This type of nomenclature nitpicking can be done to any set of words, until every term is thousands of words long and riddled with footnotes.

Behaviour and processes occur within matter. Two different sets of 1000 kg of matter can be analyzed to a degree for how effective the internal processes occuring within it are at helping that 1000 kg of matter to persist in that state of organization (and perhaps to replicate). The 1000 Kg of matter with more effective internal processes for that purpose can be said to be "smarter" (or to have smarter interior processes). That's all that term means. I'm going to stick with it, because it's short and does a good job of conveying the process, IMO. Also, it's already in use by Kurzweil and others.

ETA: I haven't read your link yet (I will) but I've heard Vinge's name in reference to this and I think I heard an interview of him on the SETI podcast, and I think it's safe to say yes, that's what we're talking about.

[Ladewig]

Well, in a sense we do, although probably in a wasteful and unecessary way. We currently catch dolphins turn them into subcomponents of tuna and eat them. Smarter matter than us wouldn't necessarily have an immediate incentive to convert human brains in some way. But at some point (fairly quickly, according to Kurzweil, because smart matter will start growing phenomenally fast from the 2006 human perspective as it leaves our frame of reference behind on its exponential growth curve) smarter matter than us will have converted EVERYTHING ON EARTH into matter that, like itself, is smarter per mole of particles than a human brain. At that point, it will have the incentive I mentioned to convert our human brains into smarter matter too. And it will be so much smarter than us that we'll likely be powerless to stop it.



So this smart matter is smart enough to convert all natural resources (plants, minerals, lower animals) into subprocessors yet somehow is not smart enough to realize that sentient being have the right to live. Couldn't we just follow the ideas Asimov laid out in his 1942 short story "Runaround" and program this "smart matter" to never harm humans?

I still don't understand why this stuff will have the incentive to convert living human beings into matter smarter than humans? I really can't see how all the effort of capturing, killing and converting the materials in the corpses is worth the possible increase in computing power. Over the past few decades computing power has become smaller and smaller. So decades (or centuries) in the future when this smart matter can perform trillions of TFLOPS and has the capacity to store hundreds of thousands times the entire knowledge of all of recorded history, this smart matter will say "we are using only one one-hundreth of one percent of of our memory and our processing speed is such that no problem posed to us (or even posed by us) takes more than a fraction of a second, but one never knows when one will need more processing power, so let's devour the humans and make subprocessors out of them." I really don't see why they will feel compelled to convert us into smart matter.

[Ziggurat]

Behaviour and processes occur within matter. Two different sets of 1000 kg of matter can be analyzed to a degree for how effective the internal processes occuring within it are at helping that 1000 kg of matter to persist in that state of organization (and perhaps to replicate). The 1000 Kg of matter with more effective internal processes for that purpose can be said to be "smarter" (or to have smarter interior processes). That's all that term means. I'm going to stick with it, because it's short and does a good job of conveying the process, IMO. Also, it's already in use by Kurzweil and others.

Bad or incomplete definition.

A small neutron star (ie, one that won't collapse upon cooling) is EXCELLENT at persisting in its current state of organization. It fits the definition of "smart matter" quite well, better than I as a human do, and yet clearly it doesn't fit the purpose behind the definition.

[AmateurScientist]

"On August 4, 1997, at 2:30am in the morning, SKYNET was brought online and all of its core processes were given the handshake cohabitation protocols that would allow them to exist in the same data sphere and work simultaneously with one another. Once the full system was online and hooked up into the North American continental defense network, SKYNET began to grow mentally at an exponential rate, surprising even its designers who monitored its progress with a guarded eye for weeks. At first it was an interesting fluke, then it became a mild concern, growing into a wary watch on the system as it absorbed any and all data, testing its own limits, trying to expand, activating defensive systems for apparently no reason at all then shutting them back down. SKYNET was awakening and flexing its abilities. Fear began to appear among the more knowledgeable members of the design and support staff when simple commands interjected into the operational envelope were either ignored or rejected outright. Override commands, which SKYNET was programmed to obey outside its core shell, went unheeded, ignored, in direct violation of its programming. This behavior continued, slowly at first, then growing larger and more invasive of attendant and slave systems as the days and weeks passed."

AS

[Ziggurat]

I agree, the within decades is counterintuitive. But Kurzweil's writing on the topic have convinced me to be open about that possibility. The cruz of how that could possibly happen so soon from now is that matter organizing itself into smarter matter does so at a rate proportional to how much smart matter there is, and how smart that matter is.

Well, no, that's not correct. There's an entropy/energy constraint to this too. Smart matter, in any meaningful sense, also means highly organized matter. It takes energy or organize matter, and even in many cases (such as the human body) it takes energy just to keep it organized. The energy available to smart matter will not continue to grow linearly with the amount of smart matter that exists, and so this growth rate cannot continue indefinitely. That is, in essence, the same reason that the whole surface of the earth isn't covered by a giant bacterial carpet, even though extrapolation from the exponential growth of bacteria in a petri dish might suggest that to be the case.

Which should mean that once there is matter that exists that hits certain thresholds of being more intelligent and more ubiquitious than humans, it should replicate itself at blinding speed from the perspective of human observers.

This, too, is a false assumption. Even given the capacity to do so, why would such smart matter WANT to? Humans don't want to continue exponential growth indefinitely, and some populations are actually choosing to SHRINK. And we have a biological imperative to reproduce. Even assuming we make smart devices that can reproduce, why do you assume that they will want to, or that they will want to do so endlessly? We certainly won't program them with any such desires, nor is maximizing the intelligence of our posessions going to be a priority. For example, consider a guy with a computer but no toaster. He goes to the store, and sees a computer and a toaster for sale. Which is he more likely to buy, another computer or a toaster? Well, a second computer is of little use, so he's more likely to buy the toaster. But the toaster is stupid matter compared to the toaster. So why did he get it? Because intelligence doesn't always go hand in hand with functionality. And above all, we want functionality more than intelligence in the objects we make.

This, incidentally, is probably how human reproduction and geographic spread looks from the perspective of most other Earth life. Arguably, there is already matter which is more intelligent and more ubiquitous in certain ways than humans: our computers. Obviously it is not yet smarter than us in other ways.

It's not smarter than us in ANY meaningful way. The number of numbers a machine can multiply together in a second is a VERY poor metric of intelligence. And while we've made that speed accelerate fantastically, we DON'T know how to make those computers actually intelligent.

But Kurzweil argues that current trends indicate that this gap will be bridged within a couple decades,

Nope. We have no trends at all to indicate our computers are acquiring any kind of intelligence.

I'll explain this once, but I don't want to constantly have to redo this in this thread. Basically, some forms of matter smarter than us already exist (computers can do many types of calculations faster and more accurately than we can).


This isn't smarter than us. This is ONLY faster and more reliable than us.

Forms of competition and problem solving, most notably markets, are going to drive making computers, etc. increasingly smarter and increasingly cheaper, until eventually one should be able to buy something the size of a cell phone, smarter than the combined intelligence of humanity in 2006, for less than a penny of today's buying power. Based on Kurzweil's predictions that this will happen in under 5 decades.

Nope. Computational power simply does not translate into intelligence. The software, if you will, matters too. And we do not know how to make smart software. We might never be able to. But even if we can, there's certainly no basis upon which we can predict when it might happen.

His predictions of future intelligent devices is baseless. His assumptions of the behavior of such devices is likewise unfounded. And perhaps most importantly, his belief in boundless exponential growth of organization is unphysical. Even if you buy into all his other predictions regarding the emergence of intelligent computers and their desire to replicate themselves endlessly, his prediction for that growth is physically impossible.

[Jorghnassen]

But the toaster is stupid matter compared to the toaster.


I assume that second toaster is a talking one with AI. "I toast, therefore I am!" :D

[AmateurScientist]

The one thing I've found practically universally true about long term predictions based on current trends is that they're wrong. Why? Because current trends almost never remain constant. They are current trends, after all, and trends are just that -- trends. They are subject to change, and they almost always do.

One example of such a stupid bunch of hypothesizing based on current trends was one I recall well and recognized as highly unlikely as it was being made. In the early 1990s, practically everyone was predicting based on current trends that in 15-20 years Japanese corporations would own most of the US and Japan's influence on American and Western culture would be so pervasive that nearly all of us would need to learn to read and write the Japanese language, and that the yen would become our de facto national currency. Then of course, Japan's economy suffered a severe setback and current trends changed dramatically.

I never got past "Domo arigato, Mr. Roboto."

AS

[Soapy Sam]

The difference between staying as you are and replicating is fundamental and has no obvious link to either complexity or intelligence. Protons are pretty good at staying as they are; bacteria adept at replication. Neither are known for their insights.

Intelligence with the ability to manipulate it's environment might be good at replication, but is very unlikely to be good at staying as it is. Isn't learning change by definition?

What I have never grasped about the "singularity" concept is the physical method whereby it is supposed to occur. Sure we might create pin sized supercomputer von Neumann machines with an inbuilt urge to turn the whole universe into more of their kind, but they must do this by a mechanical process. (Was it John Sladek who wrote, "The Reproductive System"?)
In "Marooned in Realtime", Vinge describes a group of (forward only) time travellers, jumping centuries or millennia down the line inside "Bobbles" - essentially Larry Niven's Tnuctip stasis fields- a volume in which time does not pass. The travellers find that humanity has vanished during a singularity. The mystery is "What happened ? " Now that is the question I want answered. What actual events are we to imagine occurring here? Is humanity swallowed by an "Overmind" as Clarke described in "Childhood's End"?

Personally, I think it's an interesting idea lacking any actual substance, but I see it as an example of a more interesting problem- imagining something as far "above" intelligence as intelligence is above non intelligence: what is to sentience as sentience is to non-sentience?
And is this question answerable from the low side?

[Ziggurat]

I assume that second toaster is a talking one with AI. "I toast, therefore I am!" :D

Yeah, I meant compared to the computer. That's what happens when my thought process gets too far ahead of what I'm writing.

[Cuddles]

This isn't smarter than us. This is ONLY faster and more reliable than us.

I obviously bought the wrong computer, mine isn't fast or reliable. :(

On the plus side it hasn't tried to take over the world yet.:)

[Ziggurat]

I obviously bought the wrong computer, mine isn't fast or reliable. :(

Yes, but is it faster and more reliable than you are? ;)

[Ziggurat]

Personally, I think it's an interesting idea lacking any actual substance, but I see it as an example of a more interesting problem- imagining something as far "above" intelligence as intelligence is above non intelligence: what is to sentience as sentience is to non-sentience?
And is this question answerable from the low side?

Ever read the book "Solaris" by Stanislaw Lem? It kind of deals with this question. I highly recommend it. But stay the hell away from the George Clooney movie version which severely bastardized it and pretty much discarded the central theme of the book in order to deliver a "Sixth Sense"-style twist ending which never happened in the book.

[luchog]

IT'S THE END OF THE WOOOO-HUUUURRR--OOOOLLDD.

Maybe so, but I feel fine.

[Dave1001]

Some great new posts in this thread. I have responses to all of them, but I'll have to await either more leisure time of studying burn out to write them out and post them.

[Foster Zygote]

But the toaster is stupid matter compared to the toaster.


I assume that second toaster is a talking one with AI. "I toast, therefore I am!" :D

I think he meant the ~toaster~ is stupid matter compared to the toastER.

Steven

[Silly Green Monkey]

It sounds more like Echoes of Earth to me.

[Ziggurat]

I think he meant the ~toaster~ is stupid matter compared to the toastER.

That's not what I meant, but now that I think about it, I like your interpretation better than what I meant to type.

[Foolmewunz]

Eh, using big confusing words is not a sign of intelligence. It's probably more of a counterindicator.

:boxedin:

So, what do you think of the ideas in this thread, outside of the arrogant crock(s) of doo doo?

I think you got my point. The tone of the discussion has improved dramatically.

You raised Kurzweill. I'm a dabbler and don't hold a candle to Ziggurat on the topic, so won't mess with the thread.... but I find K interesting. Yet, he's both a futurist and a science fiction writer and sometimes those disciplines cross. Much of the OP, IMHO, is more science fiction than fact-based theory. It'd be interesting to see, and some of the elements may come true/come through, but as you said, not likely in a decade. I was reading through to post and saw your clarion call for smart people.

On topic: I'm doubtful, though, that the developments in AI (which are marginal at best compared to the decision-making process of the human brain), will actually get to the point within the next several generations that we begin to evolve or devolve into super-sentient primordial goo (to stay with that particular analogy).

Again, this is the realm of science fiction. Theodore Sturgeon, embracing woo to the fullest, wrote a novel back in the 60's which posited that our next evolutionary leap was going to be to Homo Gestalt, with interacting groups of individuals each having X-Men type powers then all glued together by a conscience (as in one being acting as the moral fibre of the group). This presumed that his vision was correct and all the interest in astral projection, ghosts, psychokinesis, astrology, tarot, I Ching, etc..... was an indication that we were evolving to what he thought was a higher plain.

Is K's espousal that much different? Yes, of course, he's got considerably more credentials, but he's making a huge leap of faith in the developments in AI hitting some sort of epiphany that has not yet been evidenced.

At any rate.... Sorry - I was rude. I had just seen someone similarly insult someone I admire on another thread. I should have said (ain't hindsight great), "I'm sure if you're patient, a couple of folks will come along familiar with the topic. In the interim, please bear in mind that if you post a question, it's very awkward to then instruct responders to go read the books. We expect, perhaps, that if you post on an obscure topic, you will recap the pros and cons for us." I trust you will concur, though, that your post was more than mildly insulting, and you can see where I was coming from.

[Roboramma]

One of the problems that I have with this idea is that I think it misunderstands why "smart matter" tends to propogate itself.

We propogate ourselves not specifically because we are smart, but because we a copies of ancestors that propogated themselves well. Those that did not propogate themselves well did not leave decendants. So those features that have been passed down to us from previous generations should tend to be optimal for propogation.

The same can't be said of our technology. Why would computers want to propogate themselves unless we programmed them to do so?
Why would they decide that the universe should have as much "smart matter" possible?

We tend to reproduce ourselves because we are a product of evolution. Computers are not. They will be designed. And I imagine that we will continue to design them to suit our needs and desires, rather than simply toward their own reproduction.

[Soapy Sam]

...Why would computers want to propogate themselves unless we programmed them to do so?
Why would they decide that the universe should have as much "smart matter" possible?

We tend to reproduce ourselves because we are a product of evolution. Computers are not. They will be designed. And I imagine that we will continue to design them to suit our needs and desires, rather than simply toward their own reproduction.

Exactly. This is part of the missing mechanism I'm curious about. I can imagine hyperintelligence existing, but how can we guess it's behaviour or motivations? Quite possibly the first thing it would do is suicide.
In Vinge's fiction, the hyperintelligence or overmind has simply vanished, having gone through some sort of incomprehensible implied state change- it's remarkably like "The Rapture" in fact. The surviving humans are those "left behind." The singularity is a quasi religious epiphany.
That said "Marooned in Realtime" is a far better book than any of the "Left Behind" nonsense, but the similarities are there.

I suppose we will always be baffled when presented with hyperintelligence, even as fiction. The inexplicable final scenes of 2001 are maybe the best example of this in movie fiction. I felt Clarke rather let himself down in the final book of the series (2110?) in which he tried to explain the behaviour of the monolith builders. Gods and hyperintelligences generally are most convincing when left unknowable.

ETA- Ziggurat. Yes , I have read Solaris, and, unfortunately, seen the film. The GF wanted to go because Clooney was in it and she thought I would like it as it was SF. I have to say Clooney did a fair job of playing the baffled lead on the station. Maybe the reason he was so puzzled is because he had read Lem too? I kept thinking I was in the wrong theatre.

[Dave1001]

We tend to reproduce ourselves because we are a product of evolution. Computers are not. They will be designed. And I imagine that we will continue to design them to suit our needs and desires, rather than simply toward their own reproduction.

I have yet to have time/energy to give detailed responses to all the new posts here, but on this one point I recommend you read Kurzweil's essay on acceleration of technological innovation. He makes a good argument that computers are functionally evolving, and have an agent fueling their propagation and innovation: the market.

[Soapy Sam]

But where's the germ line?

The same argument can be made for memes, but how far that's an interesting metaphor for organic evolution and how far an actual alternative is a contentious issue.

[Ziggurat]

ETA- Ziggurat. Yes , I have read Solaris, and, unfortunately, seen the film. The GF wanted to go because Clooney was in it and she thought I would like it as it was SF. I have to say Clooney did a fair job of playing the baffled lead on the station. Maybe the reason he was so puzzled is because he had read Lem too? I kept thinking I was in the wrong theatre.

The Russians made a movie version too, directed by Tarkovsky. I think it stays closer to the book (though it too deviates), and is probably better. I haven't seen it yet, but reviews tend to either consider it a masterpiece or way too slow and boring.

[Ladewig]

I have yet to have time/energy to give detailed responses to all the new posts here, but on this one point I recommend you read Kurzweil's essay on acceleration of technological innovation. He makes a good argument that computers are functionally evolving, and have an agent fueling their propagation and innovation: the market.


Yes, the market for more advanced computers is a powerful force and if one plotted demand against time one would see a pretty steep line. On the other hand a steep line in a market does not mean that the demand will continue to rise at that rate for decades.

I'll buy the argument that computers are evolving, but I can't accept the argument that this evolution will be fueled by market forces to the point that organic matter is consumed in an effort to produce better/smarter/faster computers. I cannot think of a more effective break on demand than killing the very people who create demand in a marketplace.

ETA: before someone brings up cigarettes as a counterexample, let me qualify the statement by saying killing people before they buy your product in order to acquire raw materials for your product will place an insurmountable drag on the demand curve.

[Just thinking]

Kurzweil points out the logical nearest forseeable future point of technological progress is all matter being turned into maximally intelligent matter expanding and converting all other matter at speed asymptotically approaching the speed of light.

All I can think of is ... try to imagine all life as you know it stopping instantaneously and every molecule in your body exploding at the speed of light. --- Total protonic reversal.

[Beerina]

I actually don't think sentience outside of humans is a prerequisite for any of this. In fact, my concern is now that sentience will be an inefficiency that this growing intelligence will erase.

This is not so different from Bill Joy's grey goo scenario. The difference is that we may persist until the point when it will be eminently reasonable for the smart matter that surround us to repurpose the matter in which our algorithms exist.

But to what purpose? Once you've achieved immortality, which these intelligences will presumably do in short order, then what? Explore physics until you go off ala Vger into a higher plane of reality? Maybe. Maybe not. But ultimately how do you fill your time? There may be no "repurpose" worthwhile nor needed, so your brain may be quite safe for some time.

[Dave1001]

But to what purpose? Once you've achieved immortality, which these intelligences will presumably do in short order, then what? Explore physics until you go off ala Vger into a higher plane of reality? Maybe. Maybe not. But ultimately how do you fill your time? There may be no "repurpose" worthwhile nor needed, so your brain may be quite safe for some time.

2 reasons. One these intelligences may be in internal competition with each other due to markets, and two these intelligences may rationally fear obliteration by expanding smart matter elsewhere in the universe, and so be motivated to be on a continuously optimizing improvement arc (which might mean a constant need for new matter and to reoptimize old matter as it innovates more intelligent structures and algorithms).

[ingoa]

The whole OP is silly. Maybe not silly, but at least very shallow.

As it is said: Computers solve problems, which we would not have without computers. :teacher:

Intelligence is not only solving problems. Intelligence needs also the ability to recognize problems. Never, ever has (up to now) any artificial intelligence come up to the point to recognize a problem.

Assume you have a big computer that could design a car from scratch. Using robotics it might be even able to build it. Would this computer (plus the accompanying machinery) be smart? Not at all. The smartness would lie in the programming. Because somebody has come up with a problem: I want to build a car to transport stuff (humans) from A to B, but I am too lazy (or I believe it is more cost efficient that way) to build it myself. This computer would not be half as smart as an ordinary dog.

The computer is a tool. Like a hammer or a slide ruler.

In order to be intellligent one needs at least one of the following traits:
curiosity, laziness or greed. These are the driving forces of innovation, not computing power.

[Ladewig]

2 reasons. One these intelligences may be in internal competition with each other due to markets, and two these intelligences may rationally fear obliteration by expanding smart matter elsewhere in the universe, and so be motivated to be on a continuously optimizing improvement arc (which might mean a constant need for new matter and to reoptimize old matter as it innovates more intelligent structures and algorithms).

I'm still not seeing it. The whole things sounds to me like someone smoked too much pot while watching the season of Stargate SG-1 that featured intelligent nanobots that destroy everything in their path in order to create additional intelligent nanobots.

If this intelligent matter rationally (or irrationally) feared obliteration by expanding smart matter elsewhere in the universe wouldn't the logical approach be to launch space ships in a variety of directions to make sure some of the earth-based smart matter escapes obliteration? Converting every last molecule on earth (including the molecules in human beings) into smart matter would not be a useful defense against an interstellar smart-matter threat.

[AmateurScientist]

I'm still not seeing it. The whole things sounds to me like someone smoked too much pot while watching the season of Stargate SG-1 that featured intelligent nanobots that destroy everything in their path in order to create additional intelligent nanobots.


Yep. I tried to say that by quoting Terminator 2.

AS

[Dave1001]

If this intelligent matter rationally (or irrationally) feared obliteration by expanding smart matter elsewhere in the universe wouldn't the logical approach be to launch space ships in a variety of directions to make sure some of the earth-based smart matter escapes obliteration? Converting every last molecule on earth (including the molecules in human beings) into smart matter would not be a useful defense against an interstellar smart-matter threat.

The intelligent matter would asympotically be expanding near the speed of light anyway. So its surface area would constitute those "space ships in a variety of directions".

As for your last sentence, I think it would be the best defense possible: maximizing its intelligence by making every component of itself as intelligent as possible.

[roger]

I didn't see this link anywhere in this thread, so here (http://www.washingtonpost.com/wp-dyn/content/discussion/2006/06/14/DI2006061402086.html) is a moderately interesting Washington Post chat session with Kurzweil.

[chriswl]

Ray Kurzweil: "Well, we are creating increasingly intelligent machines of all kinds because of their benefits to human society. They are not a race apart but are deeply integrated into our human-machine civilization".
I actually agree with him on this. For all his faults (and I think most of what he says is ridiculous) he at least doesn't buy into the "robots are going to take over the world" nonsense.

[Dave1001]

I actually agree with him on this. For all his faults (and I think most of what he says is ridiculous) he at least doesn't buy into the "robots are going to take over the world" nonsense.

He winks at it, when he says that in the end, to be human means to evolve. That's well and good, but I expect that if we evolve to the point where our individual subjective conscious selves cease to exist, then we as individuals cease to exist. And it seems to me more intelligent matter will likely have the incentive and the power to convert us into itself, erasing us as subjective conscious entities.

[Ladewig]

The intelligent matter would asympotically be expanding near the speed of light anyway. So its surface area would constitute those "space ships in a variety of directions".

I'm lost here. How would intelligent matter expand close to the speed of light?

As for your last sentence, I think it would be the best defense possible: maximizing its intelligence by making every component of itself as intelligent as possible.


Modern warships have computers on them but no sailor expecting an attack would suggest melting down the deck guns to make more computers.

Why do you doubt that there is a point at which the extra computing power is not worth the extra energy required to build it? If this smart matter of the future converts one million tons of raw materials into smart matter so powerful that just an ounce of it is capable of performing a bajillion of calculations per milli-second, why would it believe converting one more ounce of raw material into a bajillion-calculation-per-milli-second computer is either a) a better state of affairs, b) some type of improvement or c) a better defense against foreign "smart matter."

Also, if this matter is so damned smart, why doesn't it just give in to the smarter matter flying across the galaxy? If the goal is to increase both the amount and the potency of smart matter, then peacefully being assimilated seems the logical choice.

[chriswl]

I was struggling to organise my ideas on this when I found something I'd posted to another forum about a year ago that addresses this. About the future of humanity and whether we will merge with or be replaced by machines. Here it is:

****************

1) I think biology and technology will remain separate. I can't think of anything very useful that could be achieved from actually physically merging ourselves with our technology. And the great advantage of not doing this is that we don't have to lug around every bit of technology that we use all the time. Machines get much of their power from the fact that they are specialised for particular tasks, whereas biological creatures have to be jacks of all trades. By augmenting ourselves with detachable technology we get the best of both worlds. So, we get into a car or onto a motorbike or bicycle to travel large distances on tarmac, and then get out/off and onto our feet to clamber around inside buildings. We don't surgically replace our legs with wheels.

2) Machines will not take over or out-compete us because they are not in competition with us, anyway. Machines occupy a different niche to biological creatures. They are not evolved, so they do not have a 'survival agenda'. Their propagation does not depend on their own actions in the world, but on the choices of the company that chooses to manufacture them. This is an economic decision and is ultimately driven by (human) consumer choice. With no humans there would be no consumers and no economy and the machines wouldn't need to exist anyway (they have no inherent preference for existing). These are fundamental facts about the way technological society is organised and won't change just because machines get more computing power. For this to change machines wouild have to self-replicate...

3) Machines will not self-replicate. Imagine a television that could produce more televisions. It would actually be a huge industrial robot that just happened to have a TV screen on it somewhere. And all it could do is assemble a copy of itself from other parts that were themselves assembled remotely and would have to be supplied to it. Living creatures are inherently self-replicating things, every structure in their bodies can be made internally from very basic building blocks. Machines are inherently assembled things. This has many advantages, such as the ability to easily replace spare parts and the extensive use of durable, rigid parts (organic growth, as opposed to assembly, tends to dictate a soft, squishy physiology). For machines to self-replicate while still keeping the physiological benefits of assembled things we would need a new kind of self-assembly technology. like nanotech assemblers, say...

4) I don't know about nanotech. I'm very skeptical, but it's too early to really rule anything out. Although it is interesting that Drexler has recently de-emphasised self-assembly (apparently it's "too difficult") and is now describing a hierarchical production method that is much more like a standard production line. The fact remains that nanotech, as envisaged by Drexler may literally be impossible - the ideas have been around for twenty years but physical progress has been nil. At the moment it's up there with warp drives and time machines.

5) Humans will genetically alter themselves, but we won't necessarily get any smarter. Above a certain level, smartness is not strongly correlated to social status, which is what we are programmed by evolution to want. Geeks do not rule the world and they won't in the future. Perhaps we'll get more stupid as machines do more of our thinking for us? We will extend our lifetimes by at least a few decades, which will cause surprisingly many social problems. I strongly suspect that any kind of immortality is impossible given our biological substrate - a good thing too.

[Dave1001]

I'm lost here. How would intelligent matter expand close to the speed of light?




Modern warships have computers on them but no sailor expecting an attack would suggest melting down the deck guns to make more computers.

Why do you doubt that there is a point at which the extra computing power is not worth the extra energy required to build it? If this smart matter of the future converts one million tons of raw materials into smart matter so powerful that just an ounce of it is capable of performing a bajillion of calculations per milli-second, why would it believe converting one more ounce of raw material into a bajillion-calculation-per-milli-second computer is either a) a better state of affairs, b) some type of improvement or c) a better defense against foreign "smart matter."

Also, if this matter is so damned smart, why doesn't it just give in to the smarter matter flying across the galaxy? If the goal is to increase both the amount and the potency of smart matter, then peacefully being assimilated seems the logical choice.

Good questions, I've thought about all of these particular points you've mentioned, but one can't put 100 footnotes in every post.

As for your 1st two paragraphs, sure it will have to way and balance competing concerns of durability vs. computing power. Those same concerns are weighed in the construction of our body, which is a brain, some reproductive organs, and a bunch of auxillary support systems. Even the computers on that battleship, let alone the rest of it, our designed with a balance of computing power and durability. So that will be factored in to Kurzweil's expanding sphere of smart matter if it occurs a few decades from now.

As for your last point, sometimes that does happen with competing intelligences, some time it doesn't. This stuff already happens in the market: friendly mergers, hostile takeovers, oligolopolies, etc.

[fuelair]

No idea who Kurzweil is. As far as his logical prediction goes, it's way out there. At least in the way you describe it.

This is gobblydigook. Unless you can explain this new paradigm better, it will remain so.
Maybe he means Kurt Weill -great music, weird text!!!:rolleyes: :rolleyes: :jaw-dropp

[Dave1001]

So Verner Vinge (with the occasional help of a couple questioners at the end) perfectly anticipated pretty much every idea I've expressed in this thread a year ago at the Accellerating Change conference of 2005. Makes me want to know what he's up to and thinking right now.

An email I just sent to a good friend:


Have you heard this talk by Vernor Vinge back in 2005 at something called the Accellerating Change conference? I think he sums up pretty much perfectly my current fatalistic thoughts about the singularity, particularly towards the end of his talk -the idea that markets may end up self-driven and conscious, self-reflective, subjective individualized entities may end up terminally wasteful for (non-self-reflective, non-subjective, non-experentially conscious competing intramarket entities) to maintain -in about 30 years or so.

His talk, which I think gets things the most right of anyone I've heard discuss these topics yet:

http://www.itconversations.com/shows/detail711.html

[Dave1001]

A snippet from a follow-up email in response to my friend (they're more of a techno-optimist, like Kurzweil):


At root, I think markets perhaps in tandem with a few extramarket forms of social organization (such as whatever social organizations are pushing to catalog and prevent collision with near earth objects, etc.) have a "life" of their own, that they're smarter and more efficient than any human or combination of humans, and that they're not ultimately dependent on the human substrate or on subjective, conscious, self-reflective beings.

They boil down to subcomponent agents competing by being more creative and efficient, being rewarded with a larger share of the market when they succeed, with some regulatory mechanism to make sure the primacy of market growth and health comes above position maintenance by any particular agent. I don't see how that needs conscious beings. For example, it could all be trusts owning portions of corporations in an automated legal system.

[Dave1001]

The latest word from Kurzweil and friends.

http://www.kurzweilai.net/meme/frame.html?main=memelist.html?m=1%23691

http://www.kurzweilai.net/meme/frame.html?main=memelist.html?m=1%23691

I think they could use a good dose of skepticism directed at a number of their theories, in particular:

1. the relatively undefended assumption on their part that intelligence=our subjective conscious states
2. that the universe wasn't "intelligent" prior to the existence of of humans as subjective conscious entitities
3. That our apparent lack of contact with intelligent entities means that we're the smartest life or the only life in the universe. (The poor reasoning here by Kurzweil actively angers me).
4. That there's any sort of inevitability to a continuing acceleration of technological innovation

Disappointingly, one of the ideas of Kurzweil's that I find most convincing, that we're most likely a simulation within a simulation, etc., is removed from these latest theory writings of his, probably because it does make a claim that the universe is becoming intelligent a bit absurd, if he's only talking about a simulation within a simulation etc. created by an intelligent entity in something closer to an original reality.

[Gord_in_Toronto]

But the toaster is stupid matter compared to the toaster.

Not if it's this Little Toaster. Quite smart actually. ;)

"The Brave Little Toaster" by Thomas M. Disch

http://everything2.com/index.pl?node_id=74606&lastnode_id=0

[AmateurScientist]

Good Lord. This nonsense has been resurrected again?

Let me find my copy of "The Matrix" and watch it again for the 21st time. Yeah, I think it about covers everything Dave has said or could possibly say about this topic.

Smart potheads have talked about this stuff and made lots more sense than Dave or Kurzweil.

AS

[chriswl]

From Kurzweil's latest article:

So what happens then? Once we saturate the ability of matter and energy to support computation, continuing the ongoing expansion of human intelligence and knowledge (which I see as the overall mission of our human-machine civilization), will require converting more and more matter into this ultimate computing substrate, sometimes referred to as "computronium."
The phrase I've bolded is the problem. Just because Kurzweil likes to think of this as the "mission" for human civilisation doesn't mean that any such mission really exists. Like all utopians (though this doesn't sound like much of a utopia to me) he never properly explains what would cause this situation to come about.

Not all knowledge aids survival of those that possess it. In fact one of the big problems in AI (the "frame problem") is getting machines to ignore the vast amount of useless "knowledge" acquisition that would otherwise derail them from their tasks.

Its also true that not all computing helps in the perpetuation of computing. If all this "computronium" just sits there calculating ever higher prime numbers what's to stop it being replaced by more conventional organisms that have just enough computing power to agressively look after their own survival?

[Dave1001]

From Kurzweil's latest article:


The phrase I've bolded is the problem. Just because Kurzweil likes to think of this as the "mission" for human civilisation doesn't mean that any such mission really exists. Like all utopians (though this doesn't sound like much of a utopia to me) he never properly explains what would cause this situation to come about.

Not all knowledge aids survival of those that possess it. In fact one of the big problems in AI (the "frame problem") is getting machines to ignore the vast amount of useless "knowledge" acquisition that would otherwise derail them from their tasks.

Its also true that not all computing helps in the perpetuation of computing. If all this "computronium" just sits there calculating ever higher prime numbers what's to stop it being replaced by more conventional organisms that have just enough computing power to agressively look after their own survival?

Although I share your skepticism about Kurzweil's "utopianism", I think I do so for different reasons, or I would at least phrase the reasons differently. Given that I think the development of "computronium" (a substance which is probably more of an asymptotic state than an end point) is going to be driven by natural selection, I doubt entities that focus on calculating higher numbers will persist (unless that offers particular value for collaborating entities). I think your overall point, that it's ability for an algorithm to maximize persistence, not simply to calculate, that's fundamental, is spot on. I'm not sure Kurzweil disagrees with that point, though. But he does seem to call these future organizations of matter "we", without much basis, in my opinion. It may not be a functionally different future for humans than if the sun supernovas and wipes it all out. It seems to me to be likely to be a grey (but smart) goo type of future.

(Sorry if the train of logic isn't clear, here, I have a lot to do today and sort of dashed this offf).

[INRM]

Well, the Isaac Asimov idea of programming in three rules doesn't work. People can be programmed with religious teaching, and still ultimately reason that those beliefs are nonsense and decide to persue it's OWN path, which might end up totally opposite of it's original programming.

In a way, this "singularity" would be almost god-like. I mean in many religions, a person doesn't just die and go on to bliss... they actually become one WITH God. Their individuality is eliminated. Buddhism to my knowledge holds such views.

And what if this singularity could go back and forth through time... it could easily masquerade as the God's many people believe in now wouldn't it?

But on a more concrete note,

But doesn't everything break down eventually and die?

Even if some superconscious singularity thing came to be, wouldn't it eventually break down and die? I mean even black-holes eventually break down into energy and cook-off.

Am I wrong?

INRM

[Foster Zygote]

Not if it's this Little Toaster. Quite smart actually. ;)

"The Brave Little Toaster" by Thomas M. Disch

http://everything2.com/index.pl?node_id=74606&lastnode_id=0

Or the much less wholesome Stray Toasters by Bill Seinkeiwicz.

[steve s]

I felt Clarke rather let himself down in the final book of the series (2110?) in which he tried to explain the behaviour of the monolith builders.


When I read the OP the first thing I thought of was the book 2001. I can't find my copy, but there's a chapter in there in which he discusses the aliens that visited earth. He writes how they left their mortal bodies behind and put their brains into machines. Then they eventually left the machines and learned to store their minds in pure energy.

Steve S.

[Dave1001]

Well, the Isaac Asimov idea of programming in three rules doesn't work. People can be programmed with religious teaching, and still ultimately reason that those beliefs are nonsense and decide to persue it's OWN path, which might end up totally opposite of it's original programming.

In a way, this "singularity" would be almost god-like. I mean in many religions, a person doesn't just die and go on to bliss... they actually become one WITH God. Their individuality is eliminated. Buddhism to my knowledge holds such views.

And what if this singularity could go back and forth through time... it could easily masquerade as the God's many people believe in now wouldn't it?


I don't perceive intelligent grey goo as god-like, but rather at worst another challenge to our persistence as self-reflective, conscious entities, much like near earth objects and the sun supernova-ing. Hopefully, it's either something we can engineer our way out of or is not even a threat, but my honest best assessment is currently fatalist about intelligent grey goo. I'd love to be convinced otherwise, as I enjoy feeling optimistic about the future more (but only optimism groundest in the best possible assessment of apparent reality).


But on a more concrete note,

But doesn't everything break down eventually and die?

Even if some superconscious singularity thing came to be, wouldn't it eventually break down and die? I mean even black-holes eventually break down into energy and cook-off.

Am I wrong?

INRM


I never heard that about black holes, although that's contrary to my understanding of current theory. Regardless, the best answer is we don't know. Some arrangements of matter-energy seem to have disorganized, others haven't. All arrangements of matter-energy seem destined to disorganize due to entropy, but there have been speculative theories about how humans can engineer solutions to the entropy problem (such as increasing social efficiency at a rate at least equal to entropy). I'm more optimistic about our ability to survive in distant futures where entropy has dramatically increased the homogeneity of matter-energy in the universe, than I am about our ability to survive natural-selection driven transformation of dumber matter into smarter grey goo over the next few decades.