The Best Cyprus Community

Skip to content


How Machines Learn

Feel free to talk about anything that you want.

Re: How Machines Learn

Postby Sotos » Sun Dec 24, 2017 12:54 pm

Yet someone not so good at reasoning/mathematics/logic can outperform someone with a high IQ by using instinct/feelings and emotional ability (Emotional Quotient, EQ).


Outperform in what exactly? Obviously not in anything that has to do with reasoning / mathematics / logic. Sure, when it comes to emotions humans can connect with other humans better than anything else... just like cats can connect with other cats. We don't expect that AI will provide motherly love to us!

but for decades now I've seen such hyperbolic excitement every time the Physicists/Engineers/mathematicians of our world 'create' something they think will mimic/outdo nature. However, so far, nature outshines the lot.


AI is also part of nature ... it is made with atoms and it obeys the same natural laws. If by "nature" you mean animals, then technology has already surpassed a lot of what animals can do. Flying was a dream for humans for millennia, but eventually technology reached the point where we have airplanes and rockets that can fly faster and higher than any bird. Communication via the internet is way beyond anything an animal can do on its own. There are machines which are more powerful even than the biggest dinosaur that ever existed. There have been many victories of technology over biology, and there will be many more. Today humans without technology are crippled. Try living for a week with just what nature provided and with nothing man-made! Then you will realize how great impact have in your life the creations of those Physicists/Engineers/mathematicians.
User avatar
Sotos
Leading Contributor
Leading Contributor
 
Posts: 11021
Joined: Wed Aug 17, 2005 2:50 am

Re: How Machines Learn

Postby Sotos » Sun Dec 24, 2017 1:12 pm

I think the analogy with flight is a good one. It took us a long time, but we eventually created machines that can fly. And we did so without having machines with feathers and without any flapping. A bird and a rocket are about as different as two things can get. GiG could say "but my bird can sing nice melodies, a rocket can't do that" or "but if I get really hungry I can eat a bird, I can't eat a rocket" ... yes, a rocket is not a bird but it can fly... even to other planets! And Artificial Intelligence will NOT be human and will not have all the OTHER attributes that humans have, but it will be Intelligent.
User avatar
Sotos
Leading Contributor
Leading Contributor
 
Posts: 11021
Joined: Wed Aug 17, 2005 2:50 am

Re: How Machines Learn

Postby Paphitis » Sun Dec 24, 2017 1:24 pm

humans are very resourceful and intelligent. Human development will not stop for anyone unless we have a mass extinction event.AI will just get better and better. Imagine what the potential is in 1000 years time in the year 3000AD.

A mass extinction event is inevitable. it's a case of when.

One of our most basic instincts is survival. I heard a talk from Elon Musk (Space X CEO) who basically said it as it is. Human Evolution will continue for the Human Race as a planetary earth species, to a Martian Species and eventually an inter universe species. In other words, we will eventually colonize other planets and universes and our survival will depend on this.
User avatar
Paphitis
Leading Contributor
Leading Contributor
 
Posts: 22295
Joined: Sun May 21, 2006 2:06 pm

Re: How Machines Learn

Postby Sotos » Sun Dec 24, 2017 1:49 pm

Personally I am not so sure that we will ever inhabit other planets in any great numbers. We have evolved for this specific planet, and living anywhere else would require creating some enclosed artificial environment. I think it would be much easier to prevent a mass extinction event, or to prepare so that we can continue living on earth even after such event, rather than living on a planet like Mars which is probably less habitable even when compared to an earth after a mass extinction event. Unless maybe we can find other planets which are very similar to earth... but those would be very very far so we wouldn't be able to reach them by 3000AD even if we departed today.
User avatar
Sotos
Leading Contributor
Leading Contributor
 
Posts: 11021
Joined: Wed Aug 17, 2005 2:50 am

Re: How Machines Learn

Postby GreekIslandGirl » Sun Dec 24, 2017 1:50 pm

Sotos wrote:I think the analogy with flight is a good one. It took us a long time, but we eventually created machines that can fly. And we did so without having machines with feathers and without any flapping. A bird and a rocket are about as different as two things can get. GiG could say "but my bird can sing nice melodies, a rocket can't do that" or "but if I get really hungry I can eat a bird, I can't eat a rocket" ... yes, a rocket is not a bird but it can fly... even to other planets! And Artificial Intelligence will NOT be human and will not have all the OTHER attributes that humans have, but it will be Intelligent.


Well, now we're back to semantics. I think it's appropriate to call what AI-computer/robots can do, or be made to do, 'artificial intelligence' but this will not amount to or equal human 'intelligence' - or, as you say, "be Intelligent".

To use your example further; there are many capable engineers (good at maths/logic/physics) but it took 2 special ones to actually make flying (sailing through air, to be more exact) possible with their machine. If you give any high-capability AI machine the problem to solve human flight (in an age before flying machines were invented so it doesn't have this pre-existing creative advantage or solution), I bet you it would never have come up with a flying machine, an airplane. Maybe AI could come up with something like Daedalus and Icarus' version which used pre-existing ideas (i.e. of birds' wings). This would be because AI can only be fed pre-existing structures and rules i.e ideas that had been fed into the 'circuit-brain' of the AI machine-bot. AI lacks that creativity that is necessary to move into different spheres. This is the same for so-called intelligent people (mathematicians etc). Human intelligence is something more. Limitless. It builds upon pre-existing knowledge and needs the spark of creativity to think out of the box. This is where emotions come in to play.
User avatar
GreekIslandGirl
Main Contributor
Main Contributor
 
Posts: 9070
Joined: Sat Oct 15, 2011 1:03 am

Re: How Machines Learn

Postby Paphitis » Sun Dec 24, 2017 2:00 pm

Sotos wrote:Personally I am not so sure that we will ever inhabit other planets in any great numbers. We have evolved for this specific planet, and living anywhere else would require creating some enclosed artificial environment. I think it would be much easier to prevent a mass extinction event, or to prepare so that we can continue living on earth even after such event, rather than living on a planet like Mars which is probably less habitable even when compared to an earth after a mass extinction event. Unless maybe we can find other planets which are very similar to earth... but those would be very very far so we wouldn't be able to reach them by 3000AD even if we departed today.


I think what they have in mind Sotos, is creating an artificial environment and eco-system on Mars. However, they are also saying that the universe and multiverse is so huge (beyond comprehension huge) that somewhere is a planetary system and a solar system with planets which are capable of supporting life just like earth.

And I agree with that thesis myself.

Humanity will need to go beyond mars eventually as the Sun turns into a Super Nover and consumes Earth and makes Mars uninhabitable.
User avatar
Paphitis
Leading Contributor
Leading Contributor
 
Posts: 22295
Joined: Sun May 21, 2006 2:06 pm

Re: How Machines Learn

Postby Sotos » Sun Dec 24, 2017 2:28 pm

Actually AI will probably be more able to "think outside the box". It already does in the fields of its expertise. For example:

As you probably already know, DeepMind’s A.I. named AlphaGo is now winning 3–1 against the top Go player of the last decade, Lee Sedol (update: the final result was 4–1). This is considered a great A.I. achievement because there are more legal Go positions than atoms in the universe. Therefore, a machine is not capable of computing all possible plays and pick the best one, but it has to develop some sort of intuition and strategical thinking in order to make the best move each time.

Lee Sedol said before the matches that either he would win 5–0 or 4–1. Imagine how it must feel to be a legend of Go and suddenly start playing against an opponent that beats you three times in a row. Moreover it plays moves that you cannot even understand.

In the second match between Lee Sedol and AlphaGo, we saw an historic moment when in turn number 37 the A.I. played a creative move that surprised all experts. Such a move was not about replicating other moves the machine had seen after watching many games. It was a move never played before by a strong player, and even knowing that, AlphaGo decided it was the best move to make.

Lee Sedol left the room for a couple of minutes after that move. The reaction of the commentators was also priceless: the move would have been classified as a mistake in any other situation, but this time no one dared to say if it was a mistake or if AlphaGo’s move was simply beyond the current human understanding of the game.


https://medium.com/@cristobal_esteban/m ... b500aa75c2

This is the same for so-called intelligent people (mathematicians etc).


Why "so-called"? Mathematicians, Engineers etc can be both intelligent and creative. Who are creative according you if not those people? Artists?
User avatar
Sotos
Leading Contributor
Leading Contributor
 
Posts: 11021
Joined: Wed Aug 17, 2005 2:50 am

Re: How Machines Learn

Postby Sotos » Sun Dec 24, 2017 2:45 pm

Paphitis wrote:
Sotos wrote:Personally I am not so sure that we will ever inhabit other planets in any great numbers. We have evolved for this specific planet, and living anywhere else would require creating some enclosed artificial environment. I think it would be much easier to prevent a mass extinction event, or to prepare so that we can continue living on earth even after such event, rather than living on a planet like Mars which is probably less habitable even when compared to an earth after a mass extinction event. Unless maybe we can find other planets which are very similar to earth... but those would be very very far so we wouldn't be able to reach them by 3000AD even if we departed today.


I think what they have in mind Sotos, is creating an artificial environment and eco-system on Mars. However, they are also saying that the universe and multiverse is so huge (beyond comprehension huge) that somewhere is a planetary system and a solar system with planets which are capable of supporting life just like earth.

And I agree with that thesis myself.

Humanity will need to go beyond mars eventually as the Sun turns into a Super Nover and consumes Earth and makes Mars uninhabitable.


If we can create an artificial environment and eco-system on Mars we can do the same much easier on earth so it can support life even after a mass extinction event. So I don't see any large number of people ever going to Mars. Maybe a few scientists, like those few scientists that live in Antarctica. Yes... with the universe being so huge the possibility of other planets like earth existing somewhere is high. But because the universe is so huge, finding those planets is difficult... and going to another star can take thousands, millions or even billions of years. As far as multiverse goes, that is just a theory and even if they exist there is no way of going there. We have more immediate things to worry about than the Sun expanding and consuming earth. It will take some billion years for that to happen... our species will probably not survive anywhere that long anyway.
User avatar
Sotos
Leading Contributor
Leading Contributor
 
Posts: 11021
Joined: Wed Aug 17, 2005 2:50 am

Re: How Machines Learn

Postby Paphitis » Sun Dec 24, 2017 2:57 pm

Sotos wrote:
Paphitis wrote:
Sotos wrote:Personally I am not so sure that we will ever inhabit other planets in any great numbers. We have evolved for this specific planet, and living anywhere else would require creating some enclosed artificial environment. I think it would be much easier to prevent a mass extinction event, or to prepare so that we can continue living on earth even after such event, rather than living on a planet like Mars which is probably less habitable even when compared to an earth after a mass extinction event. Unless maybe we can find other planets which are very similar to earth... but those would be very very far so we wouldn't be able to reach them by 3000AD even if we departed today.


I think what they have in mind Sotos, is creating an artificial environment and eco-system on Mars. However, they are also saying that the universe and multiverse is so huge (beyond comprehension huge) that somewhere is a planetary system and a solar system with planets which are capable of supporting life just like earth.

And I agree with that thesis myself.

Humanity will need to go beyond mars eventually as the Sun turns into a Super Nover and consumes Earth and makes Mars uninhabitable.


If we can create an artificial environment and eco-system on Mars we can do the same much easier on earth so it can support life even after a mass extinction event. So I don't see any large number of people ever going to Mars. Maybe a few scientists, like those few scientists that live in Antarctica. Yes... with the universe being so huge the possibility of other planets like earth existing somewhere is high. But because the universe is so huge, finding those planets is difficult... and going to another star can take thousands, millions or even billions of years. As far as multiverse goes, that is just a theory and even if they exist there is no way of going there. We have more immediate things to worry about than the Sun expanding and consuming earth. It will take some billion years for that to happen... our species will probably not survive anywhere that long anyway.


I don't know about that. Surviving millions of years on earth is probably not possible because of the possibility of a Huge Asteroid smashing into earth making all life extinct. But I believe humanity will go on and live in space on things such as Space Stations, other planets like Mars and further afield. If we have another 10,000 years of technological advancement from now, my mind actually boggles as to what humans would be able to achieve in that time. It isn't as far fetched as it might seem.

Right now our computers, phones, gadgets, technology is advancing year in and year out. Look at how everything has advanced in just a few years. Where will we be in 10,000 years time. AI will probably at the very least be running every aspect of our lives by that time. Average lifespan could be 1000 years.
User avatar
Paphitis
Leading Contributor
Leading Contributor
 
Posts: 22295
Joined: Sun May 21, 2006 2:06 pm

Previous

Return to General Chat

Who is online

Users browsing this forum: No registered users and 1 guest