AI and IQ: The Right Answer to the Wrong Question
Jul 14th, 2011 | By Timothy McGettigan | Category: Lead, Timothy McGettiganSo what is intelligence? What is IQ? What makes one person smarter, and thus more deserving of reward, then another. Well, as Tim points out, and according to many psychologists its a magic number. Like a gypsy’s gaze into the crystal ball, this number, derived with suitably esoteric and “unbiased” (not!) scientific instrumentation, reveals all. Or does it? And, as Tim asks, can it? Can a simple number like 42 really reveal all the secrets of the human experience, or is just (as Douglas Adams has suggested) a big joke.
In the Hitchhiker’s Guide to the Galaxy, a group of truth-seekers entreats Deep Thought, an artificially-intelligent supercomputer, to reveal the answer to the most elusive question in existence, “What is the meaning of life, the universe and everything?”
Deep Thought takes up the challenge, but warns that it will require no less than seven and a half million years to produce the answer. Given the scope of the challenge, Deep Thought’s petitioners accept the computer’s terms and leave it to their descendants to benefit from Deep Thought’s protracted ruminations. Finally, following eons of cogitation, Deep Thought stirs and announces, ominously, that the long-awaited answer is ready–but Deep Thought adds that the answer is unlikely to be a crowd-pleaser. Their patience at an end, Deep Thought’s supplicants insist that the computer unveil the monumental secret that they have waited so long and faithfully to hear. At that, Deep Thought heaves an electronic sigh and pronounces that the answer to the question of life, the universe and everything is…
…forty-two.
As Deep Thought predicted, the assemblage reacts unfavorably to such a meaningless answer. Deep Thought fires back that the answer is perfect, rather the fault lies with the question: the answer, forty-two, is unintelligible only because his patrons never really understood their original question.
Aha!
So, what does this have to do with artificial intelligence research? A lot, actually.
At this point, I should emphasize that I am a huge supporter of information technology and AI. The smarter that our technologies become, the more likely it is that we’ll be able to solve the many problems (e.g., war, famine, disease, natural disaster, pollution, energy shortages, etc.) that humanity faces.
That said, the conundrum that Deep Thought reveals (i.e., it is difficult to find the right answers if we don’t really understand our questions) is reminiscent of the challenges that AI researchers confront with regard to the definition of intelligence. AI researchers have a crystal clear vision of their ultimate goal: creating intelligent machines–just like that smart-aleck Deep Thought. The problem is that, AI researchers have at best a weak understanding the question, “What is intelligence?” For example, at present, one of the most widespread misperceptions about intelligence is that–get this!–intelligence can be represented as a number. Whereas Deep Thought contends that the answer to life, the universe and everything is forty-two, many intelligence experts are convinced that intelligence can be characterized as a single number; a.k.a., an IQ score. If I didn’t know better, I’d be sure that this was another one of Douglas Adams’ jokes.
It makes about as much sense to say that intelligence is equivalent to a score of 81, 97, 112, or 250 as it does to claim that the meaning of life, the universe and everything is 42. The difference between these assertions is that Douglas Adams was joking, whereas psychometricians are serious. Aww, c’mon, there’s got to be a camera hidden somewhere.
But, here’s the best part. In the Hitchhiker’s Guide, Deep Thought hatches a bold plan to solve the problem of “how to understand the question”; it’s another joke, but it contains a marvelous kernel of truth. Deep Thought creates the most sophisticated computer in the universe (i.e., the earth) and runs a multi-billion year program (i.e., the evolution of life on earth) through which to create organisms (i.e., humans) who develop the necessary mental faculties to understand the meaning of existence. Thus, the moral of the story is that the answer lies within. It simply takes the necessary wisdom to understand the question.
Poetry.
The only snag is that, at the very moment that humans finally achieve enlightenment, a bunch of Vogons blast the earth into cosmic dust. There’s a lesson in that, too. Humans have always relied on their brains for survival. So, now and forever, humanity has got to continue getting smarter or we’ll end up being obliterated by our problems.
The point of all this is that, if we try hard enough, humans will surely be able to figure out the meaning of life, the universe and everything–and also the key to artificial intelligence. The trick is to seek the real, hidden knowledge that lies behind otherwise meaningless numbers.
No related posts.
In the book “Consciousness and Language“John R. Searle (Searle, 2002) cites that philosophical problems have their ultimate solutions and that sole questions we should not take for granted. He always wants to investigate the structure of a question in order to check whether it bases itself on false assumptions or assemblies inadequate set of paradigms. This way, by deconstructing and dismantling, we come to the point where solving is possible. “Thought processes are shaped by conditions in the society within which they occur, in accordance with laws that the sociology of knowledge uncovers“(Boghossian, 2001). Therefore, there is no answer on all questions since some of them are structured in a such a manner that none answer would be satisfactory correspondent to truth or objectivity of facts it relates to. Thomas Kun came to these conclusions also and presented them in his works. He wrote that each and every theorist is entrapped inside his own theoretical universe, that is, he wanted to say that the paradigm creates its own reality. My questions are such as: can AI behave irrationally, can AI behave as if it feels the depth of desperation, meaningless, can IA behave as not being sober, can it be more human than humans are, ever. Is AI meant to save the image/idea of humans when we disappear. Did AI ever existed or it just appeared as it had been “real“? Can it ever surpass its functionalism and sense the urge to seek the answer for questions it was not programmed to? Here I wanted to excel the authors who tackled the discourse analysis and contemplated on the issue of social reality. Perhaps, AI misses that reality to become“ alive“. God forbid!
I’m afraid AI isn’t capable of much, apart from giving something inanimate the appearance of intelligent behaviour. In reality, an AI is no more intelligent than a spreadsheet, which is why it’s called ‘artificial’ intelligence. Usually it’s a program that selects actions randomly from a predefined range of possibilities, and will always be constrained to that range. The exception are the adaptive pattern recognition systems used in everyday mundane stuff.
However, genuine machine intelligence, or consciousness, is a different story. I believe it’s entirely possible that a system of sufficient complexity would become conscious under the right conditions, and I wouldn’t be surprised if the Internet itself is conscious at a very low level. The Internet is adaptive, and to some extent designed for self-preservation. Under the surface of glossy web pages, there’s a lot of network addressing, error correction and routing going on at several layers simultaneously.
Unlike science fiction, machine intelligence wouldn’t manifest itself as something immediately obvious, much less communicate in a way humans would recognise.
There’s also the Blue Brain project, which created a simulation of a neocortical column. Although that was for biological research and required tens of thousands of processors to simulate, that’s perhaps the closest we’ve come to machine intelligence.
HI,
As Deep Thought has just raised a doubt whether the QUESTION asked was correct or not, so might the human brain! What i m trying to say is that, if the question -’ What is the meaning of life ?’ is in itself very unspecific , then making it more specific actually increases the complexity(not the algorithmic complexity) of the question !..which in turn implies that we must have more knowledge about WHAT TO ASK or rather WHAT ALL TO ASK ! To get that knowledge we might resort again to ASKING general questions about the universe and things around , which again are unspecific ! Where does this go ?? NOWHERE right ! Its like TO ASK THE CORRECT QUESTION, WE MUST HAVE MORE SPECIFIC KNOWLEDGE , WHICH IN TURN CAN BE DISCOVERED BY ASKING THE SAME GENERAL QUESTION ! It seems like there actually is nothing called the ultimate truth which we are trying to find !! But in attempting to find it , we are making some interpretations about the universe around us and how it works !!
There’s no way around studying. I recommend intro books on AI. Check your local library.
Best.
TM