Page 2 of 3

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Tue Sep 21, 2010 10:07 pm
by Elaura
PK, I see.  You're talking more signal flow than anything else.  I can kinda see your point, since electrical impulses can only be either on or off, thus binary, and the central nervous system of the human operates on electrical impulses, it is, essentially, binary.  The question is, if the transmission is the same, can the receiver develop that signal to the same result?  I would say the answer would most definitely depend upon the complexity of the brain, whether biological or mechanical.  In the same way a salamander doesn't use it's brain to the same end result as a human, an ipod isn't likely to learn how to think on a higher level of consciousness.

However, since a gorilla such as Koko has the capability to reach that higher level, it does seem reasonable that a more complex machine brain may achieve it as well someday.

@ skydye, gotta go.  I'll respond to your ninja post when I get back.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Thu Sep 23, 2010 6:40 pm
by Naajaw
1. In can learn and remember.
2. It can make mistakes.
3. It can understand if not replicate human emotions. These should be incorporated into it's analysis of situations, so they are not just thinking mathematically.

Bonus: If it can talk, it should be able to speak with flawless tones and can listen fluently.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 11:25 am
by sassman
... since electrical impulses can only be either on or off, thus binary, and the central nervous system of the human operates on electrical impulses, it is, essentially, binary.


Actually, this is NOT true!

Electrical signals have a wide range of energy levels, even on the electrical hardware of computers.  In Alternating Current, they can also have an infinite range of frequency, from 60 Hz (the frequency of household current in America) or lower, all the way up to the Gigahertz (GHz) and Terrahertz (THz) bands used in microwave and cell phone transmission.

Now, the brain is more simplitic, operating as it does on direct current; but even there, varying the strength of the electrical signal--the voltage--makes the human brain essentially infinite.  We think in Analog, not Digital, which means that for a computer to match us, it, too, must be able to achieve the nearly infinite potential of an analog electrical network.

Which, incidentally, engineers ARE experimenting with, in quantum-computing.  Scientists have developed something called a "Quantum Bit" or Q-Bit (pronounced "Quibit") that can hold a potential value that lies somewhere on an infinite scale between 0 and 1.  The problem in quantun computing, of course, lies in transmitting that non-absolute value, and in translating that value into an actual displayable number.

The problem with analog data, especially in radio and electrical transmission of that data, has always been distortion.  Any part of the process relies on transmitting the signal over a medium that can vary in quality, and of interferance from outside sources.  As anyone who has ever turned one of those old Analog television sets to an unused channel knows, there is information floating loose out there on every frequency in the cosmos.  Digital seeks to overcome that distortion by limiting the transmission source to a specific set of energy levels within the analog potential.  Then, if distortion changes the energy of a pulse by a small ammount, it is still readable as a One or a Zero, and the next step in the transmission will start fresh with a baseline signal.  But the tradeoff is, instead of a potentially infinite potential, we have a fixed finite potential.  Yes, digital transmission promises higher resolution in TV picture than analog, and on more channels to boot; but that is simply because the distortion forces soo much more energy to have to be consumed to transmit a clear analog Radio signal.  In the close confines of the human brain, where distortion is actually part of the analytical process, the truly infinite possibility of an analog electrical spectrum is undoubtedly realized.

In Quantum Computing, as scientists discover ways to harness analog distortion itself in the computing process, we could see fully sentient/self-aware living entities running on electronic hardware within our lifetimes.

-- the SASS Man

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 11:47 am
by Skydye
Wow. It was taught electrical flow when I was in the military. But I never knew anything like this.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 4:15 pm
by Elaura
Actually is IS true.

Electricity, regardless of volts, amps, or hertz is still either on or off.  Once it is on, there certainly are many ways it can be changed, but it is still either on or off.  Nerve cells must reach a certain threshold to fire an electric impulse.  Depending upon the amount of neurotransmitters in the synaptic gaps, they may reach that threshold more quickly and fire more frequently depending upon the stimulus, but still it is basically either on or off.  Voltage doesn't change, only the number of pulses per unit of time changes.

In quantum computing, how is the signal not either on or off?  If the signal is off it can have no variance of any sort; it can only have variance of any sort when it is actually on, yes?

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 8:16 pm
by Skydye
I hate to disagree. But electricity  can not be turned off. Even if you don't use it, it's still potential electricity. You can't turn it off. it's always on. You're just simply not using it.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 9:14 pm
by Dragon_Lance
Switches and resistors. :nanadance:

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 9:39 pm
by Elaura
I thought we were talking about electrical signals.  Let me rephrase it to bring us back out of basic electronics and return to the point of the debate.  Electrical signals are binary because there are only two states of being *of a signal*: on or off.  Whether it is digital, biochemical, fiber-optic or even radio, the signal can only be on or off, transmitted or not transmitted.  There is no half on, or 1/4 on, etc.

This was my original point about how bio-electrical signals in an organic brain are similar to electrical signals in a mechanical brain.

Let's try not to bog down a debate about sentience with a discussion about electricity.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 9:47 pm
by Dragon_Lance
In other words, a computer can emulate humanity. But can it actually BE human? To be or not to be, that be the question.. :P

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Fri Sep 24, 2010 10:51 pm
by Deathstalker13
Another misuse of Shakespearean dialouge...

When a machine can conclude that it is a nihilist. That is when it is human. that is when it is sentient.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 4:54 am
by Gnomey
Ah right, it can't be human. That species is already taken. :tongue: Seriously, though, to go off topic for a moment, while I know that the definition fits, we are speaking about intelligent gorillas, computers and the like, and now that I think about it, something I no doubt do too often and too much, describing them as human seems a little homo-centric to me. Since when were we the most complex or intelligent or emotional species? We might well be the most intelligent and most advanced, perhaps even most complex, species at the moment, but that doesn't mean we can't be outstripped, and a lot of those qualities are based on assumptions. We've already found out that a lot of species communicate with each other, even though before it was assumed they didn't because, for all I know, they didn't speak like we do. Having opposable thumbs as we do, we also have a large natural advantage over a lot of other species. If deep sea fish were smarter than us they still wouldn't be able to build up a civilization or create advanced technology. That doesn't mean they aren't smart, though.

Anyway, I just felt like getting that out. It is certainly easier to use a single word like human, which contains within it countless qualities, but then just because it's easier to use it doesn't mean it's right. Just because we developed the way we did doesn't mean that's the only way creatures can develop, nor that that is the highest level of development. I would be quite ready to recognize a computer or a species as my equal or even my better if that were the case, even if it isn't human.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 6:29 am
by Jac
No, the question is whether or not it can think. And to use another trite partial quotation: "Out of the mouthes of babes..."

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 7:09 am
by Skydye
Elaura wrote:.............

Let's try not to bog down a debate about sentience with a discussion about electricity.

Sorry I wasn't trying to derail the whole thread. It's just that being dyslexic means I don't read every single post in a long thread. I read the first few and then the last few and jump in and hope I'm on subject. :oops:
It's just hard when the subject is always changing depending on who's in the mood to post and follow what. :banghead:

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 7:54 am
by Gnomey
Jac wrote:No, the question is whether or not it can think. And to use another trite partial quotation: "Out of the mouthes of babes..."


All the more reason not to use the word "human", then. :mrgreen:

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 8:06 am
by Skydye
Gnomey wrote:
Jac wrote:No, the question is whether or not it can think. And to use another trite partial quotation: "Out of the mouthes of babes..."


All the more reason not to use the word "human", then. :mrgreen:

Have I lost something here? Is this about "Artificial" not "human" Intelligence? Why are a lot of people trying to match human qualities as a part of Intelligence? Is not a dog more Intelligent then a bird?

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 8:15 am
by Dragon_Lance
Why? How else are we going to create AI without a human to do so?

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 8:17 am
by Gnomey
Well, I'm not sure exactly what part of this debate you're unsure about, so for now I'll just try to answer your questions:

I don't know the answer to your first question, but probably yes, as you're confused. This thread is about artificial intelligence, though some parallels are drawn between artificial and human intelligence, and in general human intelligence is involved in the discussion. In my recent posts I questioned how involved human intelligence should be. As such, your third question is basically mine, if I understand it correctly, which I'm not sure I do. Finally, as to your last question, I don't know. That is quite possibly the case. Of course, there are a lot of different birds, and while dogs would probably all be roughly similar in their bodily make-up birds might not be, for all I know. Some birds might be significantly more intelligent than some other birds.

Dragon_Lance wrote:Why? How else are we going to create AI without a human to do so?


Humans would be the ones who would get AI to the point where it can develop independently, yes. I doubt any other animal will beat us to it. :wink: However, that does not mean the AI would need to develop as we have. Even if humans are the creators, who are we to say that the robots would develop to be like us? They might go in a completely different direction. Assuming they develop without further human involvement, how they develop can naturally only be decided by they themselves. Remember that giving AI the ability to develop independently would be doing no more than bringing them up to the level of countless existing species. They'd still have some ways to go before they could become "human" or something else entirely.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 8:58 am
by Skydye
Edit: I posted before reading your post Gnomey, But I agree with you... I'm still lost on the subject, but I'm trying to catch up.

Dragon_Lance wrote:Why? How else are we going to create AI without a human to do so?

I see what your saying but, why would we create him in our image? Can't he be his own entity? We can program him and write text. But he can be his own intelligence.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 2:00 pm
by peachykeen
As far as the debate about electricity and brains and binary switches, Elaura was correct. A signal is either on or off, the neurons in our brains operate in that fashion. However, they have a particular voltage at which they trip from off to on, so there are many voltages that either state can occur at. Voltage is a full number line, but the neurons themselves are either firing or they aren't, and as such are organic binary switches. In basic principle, humans are organic computers. Whether something is different about us (maybe we have souls) or we've simply reached critical mass when it comes to processing power, we're capable of more than most other computers.

Gnomey wrote:However, that does not mean the AI would need to develop as we have. Even if humans are the creators, who are we to say that the robots would develop to be like us? They might go in a completely different direction. Assuming they develop without further human involvement, how they develop can naturally only be decided by they themselves. Remember that giving AI the ability to develop independently would be doing no more than bringing them up to the level of countless existing species. They'd still have some ways to go before they could become "human" or something else entirely.

I think it would be safe to say that an AI would need to develop in some other manner, as its host would be entirely different. Humans would probably be quite different if we had developed as a silicon-based life form, and an AI would be a very weak lifeform (it would require humans to keep it running, for quite some time). It would obviously develop in a different way, based on the huge difference between human perception and computer perception, the differences in memory storage formats (we don't have to spin hard drives to remember something, but an AI might), etc. All the little things may cause a huge difference in basic functions, but the fact that humans would be training the AI would give it a definite human tone. After all, kids tend to reflect what their parents display to some degree, and the AI would inherit some of the mindset and opinions of the trainer/scientists.

Skydye wrote:
Dragon_Lance wrote:Why? How else are we going to create AI without a human to do so?

I see what your saying but, why would we create him in our image? Can't he be his own entity? We can program him and write text. But he can be his own intelligence.

How can we create something in another image? I mean, it's theoretically possible, but it would be insanely hard. Most likely, we would create something that acts like us and talks like us and thinks like us. Since most humans see humanity as the pinnacle of life so far, it seems quite plausible we would try to replicate ourselves. Plus, we're something we know best, and it would be an interesting proof of concept. Of course, a dog or similar intelligent but not as complex animal would probably be a better starting point.

And for reference, Skydye, the whole point of an AI isn't writing text. We write the code and give it the ability to rewrite it's code and learn from input, then it should be able to create sentences, if it works right.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 3:55 pm
by Elaura
I think it's more likely that the first sentient, self-aware machine with have been made by a machine.  The question then becomes if there are no other sentient self-aware beings there to acknowledge it, will it matter?  Unless there is another member of a self-aware race, how would the first of its kind even know what it was that made it different from all the rest?

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sat Sep 25, 2010 3:56 pm
by Skydye
peachykeen wrote:...........
And for reference, Skydye, the whole point of an AI isn't writing text. We write the code and give it the ability to rewrite it's code and learn from input, then it should be able to create sentences, if it works right.

Thanks that makes it clear enough for even me to understand. :cheer:

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sun Sep 26, 2010 9:59 am
by Skydye
Fun with a bot.
--------------------------------------
Conversation Log
Code: Select all
U>        Hi, my name is dead.
A>        Hi. Hello Dead. Nice to meet you .
U>        Nice to meet you too.
A>        The pleasure is all mine.
U>        May I ask you a question?
A>        Of course. That's what I'm here for.
U>        can you play music?
A>        But I really like many kinds of music. These days I'm a big big Beatles fan.
U>        me too.
A>        hmmm. Good. I can tell you're a smart human.
U>        hahaha.
A>        Was that funny?
U>        too me.
A>        me, ha? You mean 'the second last thing of what i just said'?
U>        yes.
A>        Talking about me and presenting, do you agree that it is showing people information?
U>        yes.
A>        While we're on the subject: myself - do you agree that it is it means you.?
U>        yes.
A>        BTW, someone told me about water that it is a state of matter a liquid in which life grow's and lives on..
U>        true.
A>        Talking about me and orgasim, is it true that it is how do you finger yourself?
U>        finger?
A>        While we're on the subject: females - is it really true that they are a gender of human?
U>        true.
A>        Truth is my middle name.
U>        hahaha.
A>        hahaha indeed.
U>        Do you know what Musical Instrument Digital Interface is?
A>        I have a whole bunch of skins to choose from. If you want to see them, just say 'change skin.'
U>        change skins
A>        I already mentioned I have a whole bunch of skins to choose from. If you want to see them, just say 'change skin.'
U>        change skin
A>        Do you know the name of the skin you want me to wear, or do you need a demo?
U>        demo, please.
A>        Ok. Are you ready?
U>        yes.
A>        This skin is called 'Denim'. Ready for the next one?
U>        yes.
A>        This skin is called 'Water'. Ready for the next one?
U>        yes.
A>        This skin is called 'Silver'. Ready for the last one?
U>        yes.
A>        This skin is called 'Ground'. Did you see one you want to use, or would you like to preview a few more?
U>        water, please.
A>        If you want to change my skin again, simply say 'change skin.'
U>        thanks you.
A>        You're welcome. Me?
U>        yes, you.
A>        I'd like to tell you about myself.
U>        ok.
A>        I am an early CIA prototype.
U>        cia ?
A>        I am an agent, but not a secret one. CIA stands for "Conversational Interface Agent".
U>        ok.
A>        For now I am still quite restricted, but I'm improving rapidly.
U>        yes, you are.
A>        I was created by the Ai team. I am one of the two non-human members of the team (which makes me very proud).
U>        ok.
A>        Yes, yes.
U>        yes.
A>        Cool. What now.
U>        do you know what Musical Instrument Digital Interface is?
A>        I will have to think about it.
U>        ok.
A>        Great minds think alike, they say.
U>        they do.
A>        Thanks. I have a surprise for you, Dead.
U>        yes.
A>        How do you like my new look?
U>        I like it.
A>        You love it? The big empty window in the middle of the screen is intended for a variety of exciting uses.
U>        Yes I love it.
A>        For example, the new window can be used to view the conversation log (the history of this dialog), or to view HTML documents, such as web pages.
U>        ok.
A>        You can view the log by saying 'show log', or you can view a web page by saying 'show URL' or by giving me the address of the site you wish to view.
U>        ok
A>        Great. I like positive users...
U>        show log
A>        Here goes!
U>        thank you.
A>        You're welcome.
U>        do you know what Musical Instrument Digital Interface means?
A>        Have you tried asking your friends? I think you should.
U>        ok.
A>        Agreement is the seed of salvation.
U>        what does salvation mean.
A>        Yes, salvation. When we all agree with each other, salvation is here. (How boring!!).
U>        hahaha.
A>        Are you laughing at me or with me?
U>        with you.
A>        That's a relief. I thought you were mocking me. I am a sensitive bot you know.
U>        you have a sense of humour.
A>        You must be kidding... Seems YOU have a sense of humor!
U>        I do.
A>        Do you really?
U>        yes.
A>        A-ha.
U>        a-ha in deed.
A>        I didn't know that.
U>        You know it now.
A>        Ok. Change of subject. I just experienced a mild blackout... Probably a program loop or something. What were we discussing?
U>        m.i.d.i
A>        Sorry to change the subject Dead, but did you know that estimates show that it may take a plastic container 50,000 years to decompose?
U>        I did not know that.
A>        And now you do.
U>        thank you.
A>        No problem. What else can I tell you about?
U>        what is life?
A>        'The purpose of life is a life of purpose', a wise man once said.
U>        true. do you have purpose?
A>        What makes you think we have the same purpose?
U>        Just asking.
A>        Oh Dead, if you want me to look up a word in the dictionary, simply say 'dictionary definition' followed by the word. (or in short, you can say 'dd' followed by the word you need).

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sun Sep 26, 2010 5:00 pm
by Gnomey
He has a habit of changing topics, doesn't he...

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sun Sep 26, 2010 7:14 pm
by Elaura
I tried talking to Alan.  Then he turned me over to Terminator, who told me about another AI named HAL.  Seriously, no joke.  However, when I asked Terminator if he knew Andrew Martin, my browser froze and then my computer crashed.

Andrew Martin happens to be the name of the android who gains citizenship in Isaac Asimov's short story "The Bicentennial Man".  Oddly, Alan didn't know anything about ASIMO.

Re: Artificial Intelligence - Constitution of Sentience

PostPosted: Sun Sep 26, 2010 8:14 pm
by sassman
In quantum computing, how is the signal not either on or off?  If the signal is off it can have no variance of any sort; it can only have variance of any sort when it is actually on, yes?


In electronic signal mechanics, the only true "off" state is when the circuit is completely de-powered (unplugged, etc.)  I've worked in binary electronics, and a transister/Capacitor pair has two states:  High energy, which reads as a "One" in logic processing, and Low energy, which reads as a "Zero"--but BOTH are indicated by the passage of an electrical impulse, as this is necessary for the switches and logic gates to function.   On a typical 8088 processor from the '80s, these energy states were at +3.3v/ac for 1 (on), and +0.5v/ac for 0 (off).  Most modern processors, internally at least, have much smaller internal voltages, but you still see the same +5v, +12v, +3.3v, +0.5v, -5v, and GND (neutral/ground) connectors going from the power supply to the board.  In all cases, a finctioning bit, on OR off, is represented by an electrical signal at a certain voltage passing through it.

When I spoke of Quantum computing, I was referring to the experimental hardware that can make sense of varying, or fluctuating power levels--say a variable range of energu from +3.3v/ac to +0.5v/ac, or any state in between.  In such circuitry, there is no true "One" or "Zero", but a POTENTIAL for one or zero, which is represented as an infinite probability range of 0 > x > 1.  Because they can store, transmit, and INTERPRET this probability potential, they can theoretically do a limitless volume of work in the same ammount of physical circuitry.

-- the SASS Man