To add comments or start new threads please go to the full version of: Computer Smarts
PhysOrgForum Science, Physics and Technology Discussion Forums > News discussions > Technology News

hawksecho
http://www.physorg.com/news108520789.html

This depends, like so many things on what you mean as "smart". Computers can do very advanced math far superior to anything humans can do, and have done so for many decades. As far as abstract thought, introsption, even the ability to get depresed...When a computer is in need of therapy, I"l know were on the right track.
magpies
computers are a construct of the mind thus they will never pass the mind in ability unless the mind wants them 2 of course I can hardly make up my mind about what I want for breakfast let alone weather or not I want a computer to surpass my abilitys mby someday ill figure out what I really want for breakfast...

BUT if computers do pass me in inteligence hopefully they will have learned how to dance properly.
KB6
The four most likely possibilities:
1. We will stay just as we are until the sun dies (really unlikely). To ensure this would require a sort of global tyranny of enforced technological, social, cultural and intellectual stasis that I don't think I'd want to live under.
2. We will become extinct as a result of some natural or man-made catastrophe (a real possibility).
3. We will "devolve" into vestiges of our current selves as machines take over the functions we used to perform and we become something like "pets" or livestock - "couch potato world".
4. We will evolve or self-modify into beings with abilities far beyond what we possess today. We'll do this in order to keep pace with our technology; be free from our diseases and other frailties, maybe even including death (In fact, we're already taking steps in this direction through implants and prostheses, nanotech, etc.); do all those things we've long wanted to do (e.g.: truly colonize space in a significant way); and just to know what it's like to be beings with the ability to manipulate their "bodies" and minds in ways limited only by imagination and physics.

Of these scenarios I think I prefer the last one.

As for those who scoff at the whole idea: please show me one single law of physics or chemistry that could prevent this from ultimately being done. I am not a vitalist.

"Homos sapiens, the first truly free species, is about to decommision natural selection, the force that made us.... Soon we must look deep within ourselves and decide what we wish to become."
- Edward O. Wilson

http://i19.tinypic.com/6f5l3yp.jpg
phys_fan
Hardware speed is great, but solitaire on a supercomputer is still just solitaire. Writing the software which implements, or bootstraps an artificial intelligence is the hard part. Writing code for an artificial intelligence may be an np-complete problem which will never quite pan out on von neumann machines. Very talented AI researchers have been working on this for some time now.
tj
I lived through the "meat-machine" era, when advances in AI were "obvious" and "just around the corner". I'm not impressed by clock cycles, or ginormous amounts of RAM. Tell me the name of the software framework in which this miraculous creation-exceeds-the-creator event will happen. Outline the heuristic principles for me.

As Frankenstein found out, it's easy to make a monster. Hell, I've got a computer that does mindlessly stupid things every day. So perhaps the 10-armed bile-spitting beast from the smoking pit will arrive. It's probably wise to be ready for that. But intelligence with elegance? I scoff.
Kirk Twist
For all you skeptics out there:
Yes, it has proven fairly impossible to code AI under our current computing frameworks.
You are just not thinking enough about the future. While we do not have the computing power now, we will eventually. If we could model things on an atomic level then we could create anything, even a human brain. The brain is just a bunch of atoms! Don't think just because we are intelligent that it makes us unique and unduplicatable.
Just as "special earth" astronomers were proven so very wrong in the past, so too will "special human" scientists be proven wrong in the future
soundhertz
KB6 and Kirk Twist, I am with you. Humans have been evolving all along, and much faster than anything else. The evolution has crossed over from the physical to the mental/intellectual at blazing speed; and there is no reason to think it's going to stop - perhaps evolution, allowed to freely continue, is an infinite vector. We well may reach the point where we will begin to manipulate the universal construct itself. This would be far in the future, when the notion of smarter-than-human non-humans is old hat.

We are very close to truly realizing the successful merger of technology and biology. When we do this, and we will, everything changes. We head into a new age. The evolution of the universe, down to our own, has been shown to be nearly flawlessly exponential. The snowball has never stopped. The thrust of this juggernaut - our own combined synergy of potential - is more vast by magnitudes than we are. And unlike the "creation goes bad" movies, catering to fear of losing the comfortable past, we, through science discovery and application, will realize better and better living quality, as we always have.

Don't believe it? Think about living in ancient days, even 100 years ago, when pain, hardship, journeying, and extreme work, to name a few examples, were not - compared to today - yet mitigated by advances in medicine, construction and fabrication, locomotion, etc. But they were, compared to the periods before. And although we do still have cancer and other tough maladies, we have far less than we used to. But just looking at the future of medicine, when the incorporation of chemistry, genetics, nanotechnology, and robotics is realized, these last few stubborn pathogens and chemical/electrical imbalances will finally be corraled.

Machines smarter than ourselves will design machines we never could. The end of scarcity, which so many wars are fought over, will be realized. Gone will be the antiquated, polluting methods we have now. We will learn how to reverse the damage we have done.

Just give science, and our evolution, a chance. It's exponential, and we are seeing the harbingers of this future already, each and every day we read the news on physorg. It is happening now.
KB6
Imagine the response you'd get when trying to explain a modern notebook computer to a computer scientist from the mid 1950's:

"Believe me kid, I'm an expert in these matters. I work daily with the UNIVAC, the most powerful computer on earth. It uses 120 kw of power. The sort of computer you're imagining would have to consume electricity in the megawatts range. UNIVAC uses 5,600 vacuum tubes. Do you have any idea of how big a machine it would take to house the millions of tubes your fantastical computer would need? Forget a notebook; It would have to be as big as a small town! UNIVAC's clock speed is 2.25 Mhz, over a thousand times slower than what your talking about. The speed you'd need would turn your machine's cicuits to molten slag in virtually no time. And all that incredible power would still be limited by the memory requirements. Our UNIVAC uses 10 BIG magetic tape drives just to store the data it uses. Your machine would need enough tape to collapse the floor under its weight. For all these reasons, your half-baked fantasy is destined to remain just that. And what about the expense? UNIVAC costs over a million dollars. Your machine would cost billions. The whole notion is absurd!"

And that's the problem with thinking that is bound by the limits of contemporary technological ability (thinking "inside the box"). The only things that really limit technological development are the laws of physics and human imagination and initiative, not the examples of technology we have available to us now.
AV
Hi,

We will build smarter and smarter computers, capable of trillions of operations per second. There is no doubt about it.
But what about consciousness? I'm not foreseeing a software who becomes aware of it's own existence, therefore, machines will never be a threat to humans.

AV
MBlueD
Our creations have already surpassed us in some areas of operations (performing math mainly). I think that what these scientists mean by 'smart' are the skills of strategic thinking, decision making and awareness.
The three are linked - Strategic thinking provides, decision making selects, and awareness is the container environment for the two.
I believe AI is well on its way to achieve (or even surpass) our own skills of strategic thinking and decision making (I'm still getting pwned in Company Of Heroes by the expert AI). Awareness is a different matter, but not one that cannot be achieved one day. Still, it would be easy to control this smarter-than-us species. That is, unless someone gives them the ability to reprogram themselves or something....
If that happens, we'd be in serious trouble.



soundhertz
QUOTE
Still, it would be easy to control this smarter-than-us species. That is, unless someone gives them the ability to reprogram themselves or something....
If that happens, we'd be in serious trouble.



 

They will probably give it to themselves. But the nature of evil intent, or even desire for mastery, is an emotional function that has been fueled by our own biology of survival. In a future age, many problems that we have today will be on the wane. Extinct even. This is also thinking outside the box, and conjecture. But this notion is just as allowable as thinking that we'll use whatever powers we find to extinguish ourselves. But look, here we are, and more than ever. imho, humanity's "inherent evil" is overrated. Anthropormorphising AI may feel natural, but futurists see that notion as dubious at best, just as AV implied.


QUOTE (->
QUOTE
Still, it would be easy to control this smarter-than-us species. That is, unless someone gives them the ability to reprogram themselves or something....
If that happens, we'd be in serious trouble.



 

They will probably give it to themselves. But the nature of evil intent, or even desire for mastery, is an emotional function that has been fueled by our own biology of survival. In a future age, many problems that we have today will be on the wane. Extinct even. This is also thinking outside the box, and conjecture. But this notion is just as allowable as thinking that we'll use whatever powers we find to extinguish ourselves. But look, here we are, and more than ever. imho, humanity's "inherent evil" is overrated. Anthropormorphising AI may feel natural, but futurists see that notion as dubious at best, just as AV implied.


Imagine the response you'd get when trying to explain a modern notebook computer to a computer scientist from the mid 1950's: ....


KB6, beautiful. btw, as a tot I used to watch the construction of the UNIVAC plant when I lived in Blue Bell.
Clyde
You could create a computer with infinite speed and memory and you still would not have a "smart intelligent" computer. We already have computers that operate many times over faster then the human brain and store endless more memory. In the end AI is about software programming. That means us humans have to program something that recreates what is working in our brains. I suggest real AI is extreme fantasy and people are kidding themselves into thinking it is just around the corner.

To make AI I believe you first have to program the code that defines the properties of a soul/the experiencer with code that determines things that mimic purpose, survival instinct, experiences that are desirable and what is not desirable. Example I have chemicals started in the brain that give me pleasure or pain depending on an experience. I like the experience of tasting good pizza and drinking good beer but would not going to like the experience of drinking piss like Budweiser. Our bodies are pre-programmed to give us these corresponding chemical responses without us even thinking. So for this to work you would need to define the properties of the world environment laws that can be used to determine what is good and bad. (enormous coding effort to just recreate that) Next to make AI I believe you need a software program that can teach itself. That is how our brains work. We start with minimal abilities as a child and interact with the environment, learn and become who we are today.(No one has been able to even really touch this type of code with anything worth mentioning). Next you need to create code that mimics all the sensory input of sight, sound, vision and touch. Each one of those is an enormous coding task in themselves and something past the simple basic video camera. Just because you can create a lens that sees the environment that has nothing to do with the ability for it to translate that into an understanding of what it is seeing. And a computer being able to determine what is enjoyable music compared to what is just noise I think alone speaks to us humans as having a soul and there being ultimate limitations in what you can code with AI. How else do you explain the purpose of enjoyment of music without a soul with people having different tastes to music on top of it? And then there is probably a number of other things I am not thinking of that would have to be coded to mimic a human. For example the need to ask why? and question things or the million of other concepts that exist like this that make us human.

We have done some amazing things in the last hundred years with technology but don't fool yourself into thinking those same accomplishments all combined compare in anyway to a possible future accomplishment of creating real AI or our ability to even reach such heights. AI that exists today is such a joke compared to what the human intelligence can do. If you do not agree with me then you probably have not given this enough thought or understanding of the true effort and task of creating real human AI in a computer.
soundhertz
But this is not about making a 'real human' AI computer. No one said anything about a human computer. No computer needs a 'soul' to be unimaginably smart. Now you CAN say that at some point humans will have nanobots inhabiting brains, and contributing to a person's overall ability for memory/creative enhancement. But not the other way around! If indeed 'coding a soul' is even remotely possible, that is not even in the sights of the most adventurous futurists. They are just figuring out how to eventually build a machine that will figure out how to duplicate the human brain's massive parallelism and marry that to the computer's ever-quickening computational ability. Human brains are not that fast compared to computers in computing, but their massive parallelism so makes up for it that we are still on a par in some ways. But when artificial computing becomes massively parallel...the sky is nowhere near the limit.

The last thing we ever need to think about (and the first thing we never need) is how to give a computer 'Ego'.
clyde
True You have a good point that AI does not have to mimic the human to maybe be smart. But it seems to me if the AI can not mimic at least degrees of human purpose, learning and sensory input then it will not be all that intelligent. Its hard to get away from those basic principles and still end up with something smat because all three go hand and hand together and if you leave one out the others are pointless.

I would be surprised if we even see someone program something to the level of an ant intelligence in our lifetime. I have written my share of code in my lifetime and I have some idea what is involved. It takes alot of time to just write code to things you think would be simple. You need to think of a Chinese billion size army of coders and someone with an insane vision to manage all those coders to produce what is necessary to make a worth while AI.

Now enhancing a human's mental ability with nanobots or something else I can not speak for or against. But it does not seem too far fetched to imagine with todays technology augmenting a human's brain to calculate/think faster. In a way it already is done.. Everytime I pick up a calculator to use I am enhancing my thinking ability just that its not implanted in my brain yet.
philip347
No commnet.
meBigGuy
QUOTE
When a computer is in need of therapy, I"l know were on the right track.

Wow, at that point I'll figure we have arrived for quite some time.

All the talk about consciousness, intellegence, and so on, assume we have some idea what those things really are. And, why is human intelligence a noble goal? Look at the trouble it's got us in to to date.

kurzweil is pretty well known for his aggressive predictions of the "singularity". I hope he's right and I make it. But, I'm dubious.

But, I think the real future lies in an understanding of the potential of Quantum computing. I like what Seth Lloyd has to say. It makes more sense that anything I've heard to date, and he is getting results (watch the video).

http://www.edge.org/documents/life/life_index.html#lloyd

QUOTE (->
QUOTE
When a computer is in need of therapy, I"l know were on the right track.

Wow, at that point I'll figure we have arrived for quite some time.

All the talk about consciousness, intellegence, and so on, assume we have some idea what those things really are. And, why is human intelligence a noble goal? Look at the trouble it's got us in to to date.

kurzweil is pretty well known for his aggressive predictions of the "singularity". I hope he's right and I make it. But, I'm dubious.

But, I think the real future lies in an understanding of the potential of Quantum computing. I like what Seth Lloyd has to say. It makes more sense that anything I've heard to date, and he is getting results (watch the video).

http://www.edge.org/documents/life/life_index.html#lloyd

Physicist Seth Lloyd sees the universe as an information processing system in which simple systems such as atoms and molecules must necessarily give rise complex structures such as life, and life itself must give rise to even greater complexity, such as human beings, societies, and whatever comes next.


soundhertz
I'd say it's more like 'intelligence without baggage'. Humans have a lot of that. Unless we invent Data's brother's emotional chip, but we should never do that. methinks the whole biological matrix that supports emotion is very complex to duplicate, and for what? To loose them (realsmart computers) from the cage? Probably the computers themselves would advise against that lol.
lengould
I'm guesing that humans are just another (very short-lived) experimental step in natures apparent quest for the ultimate survivor. I wonder why (s)he is searching? Any century now, mankind is going to create his replacement, something like a self-programming self-replicating artificial (to us) intelligence.

Will it consider itself artificial? Do we consider ourselves "artificial dinosaurs" ?
PhysOrg scientific forums are totally dedicated to science, physics, and technology. Besides topical forums such as nanotechnology, quantum physics, silicon and III-V technology, applied physics, materials, space and others, you can also join our news and publications discussions. We also provide an off-topic forum category. If you need specific help on a scientific problem or have a question related to physics or technology, visit the PhysOrg Forums. Here you’ll find experts from various fields online every day.
To quit out of "lo-fi" mode and return to the regular forums, please click here.