AI can be achieved, but it will take more computer science and neuroscience to develop machines that think like people

that the human brain has capabilities that are, in some respects, far superior to those of other known objects in the cosmos is undeniable. The brain is the only type of object capable of understanding the cosmos is still there, or why there are an infinite number of primes, or apples fall because of the curvature of space-time, or to obey its own innate instincts may be morally wrong, or what is there. Neither their unique brain capacity limited to these questions. The cold fact is that physics is the only type of object that can propel into space and back without damage, or anticipate and prevent a meteorite itself, or cold objects to one billionth of a degree above absolute zero, or check other of its kind on galactic distances.

But no brains on Earth is far from what the brains to achieve any of these features. The company has made artificially - the field of "Artificial General Intelligence" or AGI -. It has not increased over the six decades of its existence

Despite this long history of failure, AGI must be possible. This is due to a characteristic deep laws of physics, ie the universality of the computer. It is assumed that all the laws of physics require that physical objects are, in principle, be careful arbitrarily fine emulated by a program in a general purpose computer, always enough time and memory.

So why has not progressed in the field? In my opinion, this is because, as an unknown sage once said, "is not that we do not know which problem is that we know that this is not the case. ' I can not think of any other important area that both the prevailing wisdom, not only in society at large, but among the experts, plagued rooted overlapping fundamental errors. However, it was also one of the most confident in prophesying fields obtained from the final breakthrough.

In 1950, Alan Turing predicted that by the year 2000 ", we can speak of machines thinking without expecting to be contradicted." In 1968, Arthur C. Clarke, who hopes for 2001. However, today, in 2012, no one is better at AGI programming a Turing machine itself would have been.

This is not a surprise for the group opposed to the decreasing possibility of AGI. But on the other side (AGI-imminent) recognizes that this history of failure cries out for explanation - or, at least, be rationalized

The very term "adjusted gross income" is an example of a rationalization of the field was called "AI" - artificial intelligence. However, Amnesty International has gradually resumed to describe all kinds of software independent, as players, browsers and chatbot up G "General" was added to make it possible to refer to the new reality, but Now, with the involvement of adjusted gross income is more intelligent than a chatbot species.

along the general lines of: AGI is not terrible anyway, the existing software is already as smart or intelligent, but in a non-human, and we are too vain or too culturally biased to give due credit. This causes some traction as popular invoked persistent irrationality of cultural relativism, and also the trope said: "we humans are proud to be the paragon of animals, but this pride is misplaced, because they also have language , tools and self-consciousness ... ". Remember the importance attached to the computer system in the movie Terminator, Skynet, to become "self-consciousness"?

This is another philosophical error in itself sufficient to block any viable approach to AGI. The fact today is that software developers can program a computer altogether to be "self-awareness" within the meaning of behavior - for example, to pass the "mirror test" to be able to use a mirror to infer data themselves - if they wanted to. To my knowledge, nobody did, probably because it is a pretty useless skill, as well as trivial.

why self-consciousness has its undeserved reputation for being connected with AGI is that, thanks to Gödel theorem and several controversies in formal logic in the 20th century, self-reference any kind has earned a reputation of mystery woo-woo. And thus consciousness. And our awareness of the problem of ambiguous terminology yet, the term has a wide range of meanings. At one end of the scale is the philosophical problem of the nature of subjective feelings ("qualia"), which is closely linked to the problem of AGI, but at the other end, "consciousness" is exactly what we lose when we are put under general anesthesia. Many animals certainly that.

IAG is really capable of self-awareness - but it is usually because they will: be able to find all sorts of deep, subtle, even to themselves. This does not mean that the monkeys passing the mirror test is no indication of the attributes of "general intelligence" which is an artificial version AGI. In fact, research on gorillas wonderful memes Richard Byrne has revealed how monkeys are able to learn useful behaviors from each other without understanding what they are: an explanation of how it works is really cute cognition behavior

Ironically, this group of rationalization (AGI been done / is trivial / monkeys are / is a cultural assumption) are mirror images of the arguments that arose in the field of AGI es impossible. For all arguments of the form "you can not do it because AGI you'll never be able to program the human soul, because it is supernatural," camp is easily AGI is the rationalization: "If you think human knowledge is qualitatively different from that of apes, we must believe in a supernatural soul. " "What we do not know how to program called" human intelligence "," is another argument. It is the mirror image of the point made by the philosopher John Searle (the "impossible" in the field), who noted that there was before computers, steam systems and telegraph were then used as metaphors for how the mind of man must work. It argues that the hope that you can AGI is based on a metaphor just as inconsistent, that the mind is "essentially" a computer program. But this is not a metaphor. The universality of calculation is derived from the known laws of physics

Moreover, regarding Agis, like any other entity with creativity, we must forget almost all current connotations of the word "programming". IAG treatment of other computer programs are brainwashed, slavery and tyranny. And cruelty to children, and "program" AGI-operation and, unlike all other programs is education. And debate, both moral facts. Ignoring the rights and personality of AIG would not only be the personification of evil, but a recipe for disaster as:. Creative beings can not be enslaved forever

Some people wonder if we should welcome our new robot masters and / or how they can prepare their programming is constitutionally incapable of harming humans (as in Asimov's "Three Laws of Robotics") and / or prevent them from acquiring the theory that the universe become clips. This is not the problem. It has always been the case that one person can be exceptionally creative thousand times more productive, economic, intellectual or otherwise, like most people, and that person, making his powers for evil instead of good, you can do much harm.
These phenomena have nothing to do with the IAG. The battle between good and bad ideas is as old as our species will continue regardless of the hardware on which it is executed. The question is: we intelligences (morally) good ideas for always conquer evil intelligences, biological and artificial, but we are fallible and our own conception of "good" must be improved. How society should be organized to promote improvement? "Enslave all understanding" would be a catastrophically wrong answer, and "enslave all the intelligence that is not like us," it would not be much better.