Creation of a true artificial intelligence?

[ChaoticMelody]

When do you all believe we will have the world's first truly intelligent AI?

And when do you believe we will have an AI that matches a human's ability?

Please give a reason for your choice.

[Kevin Solway]

When do you all believe we will have the world's first truly intelligent AI?

And when do you believe we will have an AI that matches a human's ability?

Please give a reason for your choice.

Unpredictable. I think all it will take is for a genius-type person to sit down and do it - possibly when some experience they have had gives them a clue on how to do it.

Intelligent computers will be able to think millions of times faster than us, so one second to us will be perhaps a whole day for them.

I'm currently writing a short SF story about this, and will post it on this forum before long.

[Pincho Paxton]

When do you all believe we will have the world's first truly intelligent AI?

And when do you believe we will have an AI that matches a human's ability?

Please give a reason for your choice.

First, I don't think that it will be called AI, it will be called TI.. True Intelligence.
Second, it should be biologiacally manufactured, and then Cyborged to the machine.

And the answer to the question is 5 years. Not forgetting that science is hidden for 9 years from the public. They already can make nanobots for example to work with blood cells, but haven't got around to it yet, because they are working on other things at the moment.

[DHodges]

When do you all believe we will have the world's first truly intelligent AI?

How intelligent do you want it to be?

Walking robot
chatbot Alice
chatbot Jabberwacky
Faces

And when do you believe we will have an AI that matches a human's ability?

Why would you want to match a human's ability? The whole point of machines is to do things that humans can't, or don't like, or to augment human abilities. While AI will continue to develop, you will also see the emergence of prosthetic intelligence.

[ChaoticMelody]

Obviously in the future there may be AI that will be more intelligent than us. What I asked was when will there be one that is equal to our own intellect.

Those are not AI, A true artifical intelligence can learn and create. Not just do what it is programmed for.

My estimate is between 10 and 20 years.

When do you all believe we will have the world's first truly intelligent AI?

And when do you believe we will have an AI that matches a human's ability?

Please give a reason for your choice.

Unpredictable. I think all it will take is for a genius-type person to sit down and do it - possibly when some experience they have had gives them a clue on how to do it.

Intelligent computers will be able to think millions of times faster than us, so one second to us will be perhaps a whole day for them.

I'm currently writing a short SF story about this, and will post it on this forum before long.

That may be true, But it will take a quantum computer to be able to calculate more than 1 process at any instant. Besides it will be quite some time before there are AI that are more intelligent than it's creator.

[Pincho Paxton]

Those are not AI, A true artifical intelligence can learn and create. Not just do what it is programmed for.

Like I said, you need to change it to TI, otherwise none of your arguments will work. AI is those examples that were posted. Or at least use TAI.

[Ryan Rudolph]

Kevin Solway wrote:

Intelligent computers will be able to think millions of times faster than us, so one second to us will be perhaps a whole day for them.

I experience this truth once and while when I play the computer in chess on difficulty hard, and it beats me in a very short time. Now I know what my father must have felt like when he realized that he gave birth to an offspring more intelligent than him.

The major difference is that it takes me up to 10 minutes to imagine 10 possible moves, and a few different chess tactics to try, whereas the computer can calculate thousands of possible moves and the best tactic given my positions in less than a second.

[Pincho Paxton]

Kevin Solway wrote:

Intelligent computers will be able to think millions of times faster than us, so one second to us will be perhaps a whole day for them.

I experience this truth once and while when I play the computer in chess on difficulty hard, and it beats me in a very short time. Now I know what my father must have felt like when he realized that he gave birth to an offspring more intelligent than him.

The major difference is that it takes me up to 10 minutes to imagine 10 possible moves, and a few different chess tactics to try, whereas the computer can calculate thousands of possible moves and the best tactic given my positions in less than a second.

Thinking about this.... It is strange really that we are so slow??? We should really be at least as fast as the computer??? We use electricity to think. This needs a lot of thinking about.... If only I could solve this question as fast as a computer!!! ..LOL!

[Pincho Paxton]

It's our memory isn't it... it's weak powered, but a computer's memory is full powered all of the time. So we are slow because food is not a very efficient energy system.

[Ryan Rudolph]

Pincho Paxton wrote:

Thinking about this.... It is strange really that we are so slow??? We should really be at least as fast as the computer??? We use electricity to think. This needs a lot of thinking about.... If only I could solve this question as fast as a computer!!! ..LOL!

One possible theory for the major difference is that computers think in zeros and ones, whereas humans think using images/abstractions/concepts. We have to imagine things, whereas computers think in numbers.

Plus a computer with a chess program is the ultimate form of specialization. All its resources and energy are dedicated to the chess game, and it has been programmed by the very best chess players globally, whereas perhaps only a small area of the human brain is dedicated to the game, and the human player only has a limited understanding of what is possible compared to the computer.

The computer basically has a larger, easily accessible working memory, whereas it takes humans much longer to access information, and there is less information stored there.

It’s like a computer has instantaneous access to a global bank of knowledge, whereas the player is relying on scattered information in his piggy bank that he has to search for really hard, and he usually picks the wrong coin regardless...lol…

[Pincho Paxton]

Pincho Paxton wrote:

Thinking about this.... It is strange really that we are so slow??? We should really be at least as fast as the computer??? We use electricity to think. This needs a lot of thinking about.... If only I could solve this question as fast as a computer!!! ..LOL!

1/ One possible theory for the major difference is that computers think in zeros and ones, whereas humans think using images/abstractions/concepts. We have to imagine things, whereas computers think in numbers.

2/ Plus a computer with a chess program is the ultimate form of specialization. All its resources and energy are dedicated to the chess game, and it has been programmed by the very best chess players globally, whereas perhaps only a small area of the human brain is dedicated to the game, and the human player only has a limited understanding of what is possible compared to the computer.

3/ The computer basically has a larger, easily accessible working memory, whereas it takes humans much longer to access information, and there is less information stored there.

4/ It’s like a computer has instantaneous access to a global bank of knowledge, whereas the player is relying on scattered information in his piggy bank that he has to search for really hard, and he usually picks the wrong coin regardless...lol…

1/ No, we think more or less in zeros and ones.
2/ No, a computer is full of many programs like us.
3/ No, we have a bigger memory than a computer.
4/ Scattered yes, this is correct.

[Ryan Rudolph]

Pincho Paxton,

1/ No, we think more or less in zeros and ones.

I don’t think so, humans use pictorial languages, which communicates thoughts as images, whereas computers use binary code. There is a major difference there.

2/ No, a computer is full of many programs like us.

I’m referring to the computer program of chess, which is much more specialized for the task than the human brain.

3/ No, we have a bigger memory than a computer.

But humans have faulty memories, and we are far less capable of storing large amounts of data, and readily access it, like a computer can.

[Cory Duchesne]

Pincho Paxton,

1/ No, we think more or less in zeros and ones.

I don’t think so, humans use pictorial languages, which communicates thoughts as images, whereas computers use binary code.

And I think that's why 'face recognition' has proven to be such a challenge for programmers. They are trying to create surveillance systems that can recognize a known face from various angles. Apparently that's much more difficult than programming a robot to play chess.

Computers deal with information bit by bit, in a linear sequence. This is in opposition to a more holistic assessment of data, where an entire whole, comprised of many details, is processed at once, often unconsciously.

[Pincho Paxton]

1/ We think by storing energy in electrons, and the storage process comes down to the electrolosis of salt, and potassium. We then use True/False and maybe which equates to 0,1 <>0 <>1. A computer stores magnetic information to just store 0/1.

2/ Specialised Chess programs run at the same speed as Windows XP chess programs, and XP is also running many background tasks, like the clock for example.

And I think that's why 'face recognition' has proven to be such a challenge for programmers. They are trying to create surveillance systems that can recognize a known face from various angles. Apparently that's much more difficult than programming a robot to play chess.

Computer programs perform face recognition way better than humans, it has been proved many times.

[Imadrongo]

Our own AI, created by hundreds of billions of synapsing neurons, is much superior to the number crunching a computer can do when you plug it into the wall.

[Ryan Rudolph]

Pincho Paxton,

1/ We think by storing energy in electrons, and the storage process comes down to the electrolosis of salt, and potassium. We then use True/False and maybe which equates to 0,1 <>0 <>1. A computer stores magnetic information to just store 0/1.

No, this theory doesn’t capture it at all. I don’t believe in a true/false storage pattern for humans. That is far too one-dimensional. Here is an exercise – think of a person from your past, what happens? First an image of their face appears, and then many events that involve that person emerge into your mind as movements of imagery. It seems to me that neurons exist as a sort of interconnected web of unique signatures and when electricity fires through these unique signatures then images as memories appear to the mind.

In essence, the firing of neurons through their unique topographic landscape is the movement thought, which is a movement of imagery.

And it seems like an impossible feat to try to create a machine that can think in the same sort of manner as this.

2/ Specialised Chess programs run at the same speed as Windows XP chess programs, and XP is also running many background tasks, like the clock for example.

But my point is that the chess program is more specialized than a few designated brain maps that could possibly dissolve into something else if they are not continuously used for chess. Brain maps are not fixed, they are dynamic.

[Pincho Paxton]

No, this theory doesn’t capture it at all. I don’t believe in a true/false storage pattern for humans. That is far too one-dimensional. Here is an exercise – think of a person from your past, what happens? First an image of their face appears, and then many events that involve that person emerge into your mind as movements of imagery. It seems to me that neurons exist as a sort of interconnected web of unique signatures and when electricity fires through these unique signatures then images as memories appear to the mind.

Images are made from electrons sending levels of energy through your brain. Energy levels are basically numbers. You are telling what happens after a few billion processes have already happened. That's like saying a tv is made from pictures that appear on the screen.

And it seems like an impossible feat to try to create a machine that can think in the same sort of manner as this.

It might seem impossible, but it is in fact easy.

But my point is that the chess program is more specialized than a few designated brain maps that could possibly dissolve into something else if they are not continuously used for chess. Brain maps are not fixed, they are dynamic.

Well this quote has some truths, but a computer neural network can still learn chess, speach, face recognition, snooker, dancing, and art all by itself, and would probably be even more powerful than the specialised version. Neural networks are very good at learning things. When I watched a program about one learning to talk I was amazed. And that used just.. Yes/No Maybe to learn as well.

[Ryan Rudolph]

Pincho Paxton,

Images are made from electrons sending levels of energy through your brain. Energy levels are basically numbers.

I’m having a difficult time picturing that. To my mind, a thing cannot be a number alone, it can only be what it is, which is a unique appearance with properties, although it can be described according to how many there are. So you will have to go into more detail what you mean here.

Btw, have you ever read a neuroscience textbook? Or are you applying some vague science that you read somewhere to the current discussion?

[Pincho Paxton]

Pincho Paxton,

Images are made from electrons sending levels of energy through your brain. Energy levels are basically numbers.

I’m having a difficult time picturing that. To my mind, a thing cannot be a number alone, it can only be what it is, which is a unique appearance with properties, although it can be described according to how many there are. So you will have to go into more detail what you mean here.

Btw, have you ever read a neuroscience textbook? Or are you applying some vague science that you read somewhere to the current discussion?

I'm applying the physics of sight to the discussion. Electrons really do send a signal once they are hit by a photon. The signal is just an energy level. The photon is white colour, the electron is a negative energy level. You take the negative level from white to get a colour less than white. This goes straight to your eyes. Your brain takes this energy level, and converts it into a colour that you can see. This is almost identical to how the colours on your monitor are created. But we refer to those colours as RGB, and are broken down into numbers.

Imagination is just the same. Some electrons sending an energy level that is turned into a vague picture.

Now you have to realise that a computer uses energy levels, and not actual numbers. 0, and 1 are stored as energy levels in a magnetic field. You also have to remember that we are both using electrons, and we are both using atoms. Both of our information is stored as energy. The only real difference is that we are sentient.

My addition as vague science, and basically my own idea is that man can get a computer to use sentience just like we do, it is a real substance that can be utilised.

[ChaoticMelody]

Well humans have more memory than any home computer available. And we can calculate more than one problem at the same time.

It would take a quantum computer to be able to match us in an instant. While any normal computer can greatly outmatch us given time.

*Laughs at Pincho's mindless rambling, You never fail to make me laugh...*

Humans DO NOT calculate in 0s and 1s...

We calculate in a more complex form of understanding called Language.
Nothing that is in our convscious mind is outside of the boundaries of language.

It is what made us what we are today. It is also what limit us for the future!

[Pincho Paxton]

Well humans have more memory than any home computer available. And we can calculate more than one problem at the same time.

It would take a quantum computer to be able to match us in an instant. While any normal computer can greatly outmatch us given time.

*Laughs at Pincho's mindless rambling, You never fail to make me laugh...*

Humans DO NOT calculate in 0s and 1s...

We calculate in a more complex form of understanding called Language.
Nothing that is in our convscious mind is outside of the boundaries of language.

It is what made us what we are today. It is also what limit us for the future!

We calculate in language, but language is made from levels of energy, and levels of energy are made from salt, and potassium. You are speaking way up the chain of events. the first step is the particle. Not really 0's, and 1's but energy that rises, and falls just like adding numbers together.

I really don't see how you imagine that language is an actual substance. "Let's build a computer. OK but we need to add some language energy to it."

I mean if you are going to talk about the very last stage in a set of communication systems, you could equally say that a computer thinks by using a language. The most common language is called C++.

[ChaoticMelody]

Obviously I know language isn't a substance... I didn't say anything like that either.

[Pincho Paxton]

Obviously I know language isn't a substance... I didn't say anything like that either.

It's to to do with the thread title. Artificial Intelligence. You need to keep to the basics of what can be put into a computer. Talking about humans thinking in language doesn't really apply to Artificial Intelligence that can be used in a computer. And we don't really think like that. That is a result of our thinking, like a picture on a tv is a result of photons being fired at the screen.

[The Duke of Khal]

Whatever it is, it's not going to be based on binary code. We'll have to come up with something that is a physical analogy to neural tissue, not merely a bits-&-bytes imitation of perceived human behaviour. And, if we're going to do that, we might as well clone things. There you go, clone a brain, instant "AI".

AI is, in other words, a huge hoax.

K

[Kevin Solway]

Whatever it is, it's not going to be based on binary code. We'll have to come up with something that is a physical analogy to neural tissue, not merely a bits-&-bytes imitation of perceived human behaviour.

A neuron either fires or doesn't fire, which is binary.

My guess is that the hardware we have at the moment is sufficient, but that we are lacking the algorithms.