#571 by igorfrankensteen » May 24, 2015 3:09 pm
I was too exhausted by the combination of stresses in my private life, the attempt by some members here to "win by way of insult," and by the excursions into areas of detailed science which I have zero knowledge of.
However, to take up where I left off...
When I referred to machines having to CONSCIOUSLY do whatever they do, I was speaking in the context of what I said in those posts, which was that everything our current definition of "machines" do, is programmed into them directly, either mechanically or through software Even when random number generators are inserted into them, in order to make them develop more varied and randomized results, they are still doing WHATEVER they do, exactly as they are originally designed to do. They do not do anything that they are NOT specifically designed to do, in the way that humans can, when they act out of UNCONSCIOUS sources.
Some people might want to imagine that "subroutines in software are akin to subconscious calculation in human minds," but I would suggest that this is only wishful thinking on their part. Because we still have no real idea how human minds work, we only have partial understanding of how SOME of the electrochemical processes involved proceed.
One of the primary things I like to do here, and in other forums I participate in, is not necessarily to directly attack a given hypothesis. What I most like to do, is to try to point out or suggest how (as is very often the case) some very important "big picture" elements to the subject are being overlooked or assumed.
That is why especially in the case of a discussion comparing machine "intelligence" to human "intelligence," I hope to draw everyone's attention to the assumptions and overlooked items inherent to THIS situation. That includes especially, not only that "human intelligence" is hard to define, it also completely misses that "surpass" as a point of decision about machine performance, directly infers that both an officially recognized authority exists to make such a judgment (a VERY dicey assumption), and even more critical, that the act of "surpassing" is possible to recognize, when and if it happens.
I look in particular, at one of the gravest threats currently facing humankind: it is a machine of sorts, and there are a lot of people already declaring that this machine has ALREADY surpassed human intelligence, and should be declared to be our "decisional" overlord.
I am referring to the various people who insist that some variation of Capitalism is "smarter" than all humans, and that whenever the machine of "capitalism" decides that something should cease to exist, or that something or someone else should knuckle under and cease struggling, that all of the rest of us should accept the "machines" decision.
No, I am assuredly NOT trying to lead this thread on a tangent for political purposes. Since the people who promote this idea set, do proclaim that their financial models, which they can and do describe via computer programs, are superior to any decisions which humans as a group, or as individuals can make, they are an exact example of what this thread is focused on. They are also an excellent illustration of the "macro" concerns that I think must be addressed, in order to rationally discuss machine intelligence, as compared to human intelligence.