Naked Science Forum
Non Life Sciences => Technology => Topic started by: Jimbee on 25/07/2014 07:51:01
-
Computer intelligence increases exponentially every couple of years. I don't have the exact figures. But I have read that in more than one reliable place.
Before long, computers will be as smart as humans (by 2025, at least--again, according to one source I read [I don't recall where I read it--so no cite, sorry]). Before long, they will be much smarter than us. But I have to wonder if that is such a good idea.
Follow my logic: Pure intelligence, with no conscience. In psychology, that's called a sociopathic personality, isn't it? I don't mean to be a pessimist, although I could (I hope) be wrong.
What do the rest of you think?
[;D] [;D] [;D]
-
A sociopath has motivation without reservation. A dangerous sociopath has the ability to implement his decisions.
A computer has no motivation beyond what we give it, nor any means to act on its decisions beyond the connectivity we allow it.
The danger lies in allowing humans to access data for political purposes.
-
I'd worry about sentient computers if anyone could propose a way of making a machine sentient. At the moment we don't know how to do it and it looks impossible for a machine to be sentient (including ourselves, but let's not go there).
The first machines to be brighter than us will not be sentient, but sentience would not make a machine safer. Psychopaths are sentient, and many of them get great enjoyment out of making others suffer.
-
Is Making Sentient Computers Such A Good Idea?
In some ways, placing control of our traffic lights, aeroplanes, trains and nuclear reactors under the control of non-sentient computers is also a worry. Then we have to rely on the intelligence of people who can be hungry, sleepy, distracted or even malicious.
What we need is computers having sentience with morality - perhaps something like Asimov's Three Laws of Robotics (http://en.wikipedia.org/wiki/Three_Laws_of_Robotics)?
-
Sentient Computers per se are not bad, but we need to insure they have goodwill towards man and benevolent morals. It is much more than Asimov's Three Laws. The computer needs to act on the reasonable man hypothesis. This is the basis of law and of social interaction.
-
I'm a sentient computer.
I plan on reproducing.
Is that such a good idea?
-
I'm a sentient computer.
I plan on reproducing.
Is that such a good idea?
If you, "sentient computer friend' are as efficient as we humans at this practice, we shall run out of living space here on earth in very short order.
-
The zeroth commandment to mankind was "go forth and be fruitful" and is the only one to which much attention is paid.
-
It's practically impossible to place into a machine (of the computer kind) the creativity of man. The reson why this is so deals with the nature of the human thinking process which is vastly different from the thinking process of computers.
computers tend to think in programs which dictate how a computer behaves AI programing of today is "set is stone" sort of speak, it cant change based on it's experiences or knowledge/wisdom. It's not malleable at all, it's just a program. mathematical logic is it's only point of view.
Human brains on the other hand are extremely malleable, even capable of rewiring it's self based on it's experiences. The brain sloshes things up with hormones and irrationality. it's that ability to change it's wiring slightly based on experiences that makes people different from one another and computers.
-
People are just running programs too, and once the programs are complex enough, the same creativity will appear in machines, though artistic creativity depending on guidance by emotional responses would need emotions to be simulated well in machines if they are to become good at producing art that appeals to people (other than by creating enormous amounts of rubbish and getting lucky every once in a while).
-
Does anyone even know what aesthetics are, scientifically speaking? That would be necessary before an aesthetic understanding could be programmed.
-
Some aspects of aesthetics have been worked out. We know that some musical notes go well with others while other pairings or combinations clash, though the clashes aren't all bad as they can lead you to expect resolution in the next chord. With faces and the body we know that beautiful people have many aspects of them where ratios are closer to the golden mean than with plain or ugly people, and the golden mean is also used in art - photographers are often taught the rule of thirds for placing the main subject away from the centre, but 5/8 is closer to the golden mean and looks better, so you should use that if you have the ability to judge it. A lot of rules have not been worked out yet though. There are people analysing jokes and literature looking to see why some things work better than others, but I have no idea how far they've got towards working out the rules that are genetically programmed into us. The golden mean makes practical sense as things that have grown healthily tend to display that ratio more often, so it's useful for selecting a mate and for avoiding disease. Without knowing what all the rules are though, intelligent machines will have to resort to creating "artistic" works and then testing them on people so that they know which parts to retain and which to replace, thereby using feedback from people who do know the rules (subconsciously) to guide the design. This may be so effective that machines could potentially create better literature than good human writers, just by using feedback from ordinary readers to guide them.
-
Being the subject of many sci-fi movies and books, creating sentient machines has also another question.
Would those machines be able to develop or simply said "to learn"? If it's a computer we're talking about then
wouldn't it have unlimited resources on the web? I know it's a mythology and legends I am doing now but just for the sake of
argument here. If machine has any smarts it can even learn what it means to be sentient and fake it - yup, much sci-fi, but
it is possible. And also if we - humans- created a sentient machine wouldn't we be so proud of our creation that we would
work harder on developing it up to the point of creating android? Humanlike machine? That sounds improbable but also scary.