Originally posted by KellyJayA blue screen is there for one purpose, and one purpose only. To tell you that the computer did something unexpected and something undesirable.
Blue screens are by design, it breaks down you an error message telling you something about the fault.
I just can't seem to reconcile the existence of blue screens with:
If they started acting outside of their function such as getting a math problem wrong they would be worthless and untrustworthy.
and other statements by various people to the effect that computers only do what we want.
1 edit
Computers do what any intelligent human being or team of human beings can do.
The advantage is they can do it usually a great deal faster.
They can do it over and over again without getting bored.
The program that runs the computer does what you tell it to do.
When there is a problem with the outcome of a program, the programmer has to go back to the program and first figure out what she TOLD the computer to do.
This takes a kind of detachment and objectivity. The machine does not do what you want it to do. It does what you TOLD it to do. And a great deal of computer program debugging is going back first to ascertain exactly what you told the computer to do.
If the machine could speak it would say "Hey, I am only doing what you TOLD me to do."
Many times the "bug" lies in the fact that the programmer did not TELL it to do exactly what he wanted it to do.
I was never a designing engineer. So I won't say much about that aspect of it.
Originally posted by jaywillI get what you are saying, and this is true most of the time. However, it is perfectly possible to take a computer to a higher level and have it do things that are more a result of environment than they are of intentional programming. This is especially true of any form of AI that has learning capabilities, but the basic concept is present in all computers.
The program that runs the computer does what you tell it to do.
Originally posted by JS357You nailed it, my friend😵
Yes, there is no knowing before there is knowing.
We can imagine a model of the world in which what is known, comes to be and persists only so long as it is known, or we can imagine a model of the world in which the known exists and persists before and during those times it is not known. I think we tend to accept the latter model as "real" except when doing philosophy. 😉
1 edit
Originally posted by twhiteheadThat's not a road I care for.
As I said earlier, we are going down the road of "they are designed to do anything so anything they do is by design".
I take your point about objects not being confined by the purpose they were designed for. A rock has not been designed yet it is fit for the purpose of cracking an egg. Yet I fail to see the arc from this aspect of being to AI. A piano is designed to make music but happens to make an excellent drink holder too. So what?
I suppose one question is the meaning of 'do' and its connection with volition. If a coconut falls from a tree and kills Sophocles, has the coconut done anything? No, it has not really done anything. Of course our language tends to personify objects - but a clock does not 'keep time', a clock is a mechanism we have designed for that purpose, it is a process unfolding set in motion by ourselves.
One problem is that our language is fit for our purposes and there is no reason an AI (or an alien, or a god) would share our language. China Mieville's book Embassytown deals with this very well, hooray for weird fiction.
I am interested in your views on AI, though - you speak about it as though it is real, not just potential, with passionate fervour - so please expound, if you would.
1 edit
Originally posted by twhiteheadAny examples of forms of AI with learning capabilities that presently exist would be helpful.
I get what you are saying, and this is true most of the time. However, it is perfectly possible to take a computer to a higher level and have it do things that are more a result of environment than they are of intentional programming. This is especially true of any form of AI that has learning capabilities, but the basic concept is present in all computers.
Incidentally, apart from the Platonism around numbers, your views are close to Daniel C. Dennett's, have you read him?
- Joined
- 31 May 06
- Moves
- 1795
1 edit
Originally posted by jaywillYou have obviously never heard of evolutionary algorithms
Computers do what any intelligent human being or team of human beings can do.
The advantage is they can do it usually a great deal faster.
They can do it over and over again without getting bored.
The program that runs the computer does what you tell it to do.
When there is a problem with the outcome of a program, the programmer has to go back to do.
I was never a designing engineer. So I won't say much about that aspect of it.
http://en.wikipedia.org/wiki/Evolutionary_algorithm
They are 'designed' by artificial evolution and it is not only possible, but usual for no human to comprehend
how they work.
As computers become more advanced and the programs they run become even more sophisticated we
will be less and less able to understand how they work.
1 edit
Originally posted by googlefudgeThe question is whether consciousness is algorithmic or not. Philosophers like Dennett think it is; philosophers like Penrose, not. Others, like McGann, think the question of consciousness unanswerable. What do you think? In my opinion an algorithm is a virtual machine, very clever mind you.
You have obviously never heard of evolutionary algorithms
http://en.wikipedia.org/wiki/Evolutionary_algorithm
They are 'designed' by artificial evolution and it is not only possible, but usual for no human to comprehend
how they work.
As computers become more advanced and the programs they run become even more sophisticated we
will be less and less able to understand how they work.
A good example of human minds not coping well with algorithms is algotraders, how they have effed things up.
1 edit
Originally posted by Bosse de NageTwo human babies are born. Identical twins. They are separated at birth, and grow up in different homes. At age thirty, what will the differences be? What will the similarities be? The differences are a result of environment. The similarities a result of DNA, or 'design' if you will.
Yet I fail to see the arc from this aspect of being to AI.
I think twins are often more similar than we would like to admit (ie we don't like to think that we are a function of our DNA). But they are never truly identical.
A computer (or computer program) that is designed for AI, can learn. What it learns, affects how it thinks. A sufficiently sophisticated AI could become concious then learn like a human, gain character like a human, maybe even get religion. But although much of the structure and programming would be designed, and possibly intentional, there would always be some elements that simply worked unintentionally, or don't work as planned, but far more importantly, the learning aspect means that the character of the conciousness will be partly a function of its environment.
If you take two identical AI computers and put one with an alcoholic father and the other with a loving mother, you will end up, with two very different computer AIs, just as you would with identical twins.
But most importantly as regards previous discussions, if you put one with a dolphin, will it end up being more like a dolphin than a human? Will it do maths like a dolphin?
Originally posted by Bosse de NageI am afraid I don't have any. I am waiting for parallel processing to become more mainstream, then I would really love to get into AI development.
Any examples of forms of AI with learning capabilities that presently exist would be helpful.
Incidentally, apart from the Platonism around numbers, your views are close to Daniel C. Dennett's, have you read him?
No I haven't, but I'll look him up.
1 edit
Originally posted by twhiteheadI'm asking for further evidence of this beyond your bare assertion. ^^Ok, you answered this. Well, we'll just have to wait and see. It's interesting, though, that your language shows you already believe AI to be real avant la lettre.
A computer (or computer program) that is designed for AI, can learn. What it learns, affects how it thinks.
Kaspar Hauser couldn't speak let alone do math.
- Joined
- 31 May 06
- Moves
- 1795
Originally posted by Bosse de NageI don't know.
The question is whether consciousness is algorithmic or not. Philosophers like Dennett think it is; philosophers like Penrose, not. Others, like McGann, think the question of consciousness unanswerable. What do you think? In my opinion an algorithm is a virtual machine, very clever mind you.
A good example of human minds not coping well with algorithms is algotraders, how they have effed things up.
I do know that the only way to find out is to do experimentation and research into how brains do work
and no amount of philosophising on the subject will get us the answer.
I also know the conciousness is a by-product of the physical activity of our brains and barring some magic
and unreproducible property of organic chemistry it must be possible to reproduce this phenomena by some
other (artificial) means.
Whether or not we would be capable of understanding how these machines work any better than we understand
how we work is a different question.
However barring someone coming up with a proof one way or another all debate on whether its possible to know
is just so much hot air.
We do science to find things out, and science keeps going until the question gets answered one way or another.
If it never gets answered, it never stops.
Saying 'it's impossible to know x', without any formal proof that this is so, is just trying to shut down the discussion and
to stop looking for answers.
It is the same thing when people talk about the 'god of the gaps' and try to explain things by saying we can never
understand because god did it.
It's just an excuse to stop, looking, to stop thinking.
However given the definition of algorithm given on Wiki... http://en.wikipedia.org/wiki/Algorithm
"... an algorithm is an effective method expressed as a finite list, of well-defined instructions, for calculating a function."
I would say that the brain doesn't appear to, or feel like, it operates algorithmically.
But I am not an expert, and I don't think we have evidence to say either way.
Which brings me back to "I don't know".
Originally posted by googlefudgeHow do you know this - in advance of the research?
I also know the conciousness is a by-product of the physical activity of our brains and barring some magic
and unreproducible property of organic chemistry it must be possible to reproduce this phenomena by some
other (artificial) means.
- Joined
- 31 May 06
- Moves
- 1795
Originally posted by Bosse de NageThe research that proves beyond any reasonable doubt that we are the product of our brains and that our brains
How do you know this - in advance of the research?
are physical is already done.
It's how I know beyond a reasonable doubt that souls don't exist and thus that there is no afterlife, good or bad.
Precisely how conciousness works is an open question that may or may not be answered.
But the question as to whether we are the product of the physical workings of our brains... done and dusted.
Answered beyond any reasonable doubt.
Of course science will constantly keep testing that, and if new evidence were to come to light that showed that
this was wrong then the scientific view would change... But given what we know already this is not a reasonable
expectation.
Originally posted by googlefudgeI'm glad it's all wrapped up for you, not a few respected philosophers, scientists and philosopher-scientists reckon consciousness remains a bit of a mystery. Your language evinces a mind made up, probably for reasons of emotional security.
The research that proves beyond any reasonable doubt that we are the product of our brains and that our brains
are physical is already done.
It's how I know beyond a reasonable doubt that souls don't exist and thus that there is no afterlife, good or bad.
Precisely how conciousness works is an open question that may or may not be answered.
But t ...[text shortened]... ic view would change... But given what we know already this is not a reasonable
expectation.