Actually, I think it would be possible for computers to be random. Not the ones now, certainly, because how could they choose such a number? They can only give output based upon input. But what if it chooses based on an external randomization say, something based upon quantum mechanics? That, if my memory serves, works upon the probability of things existing, but the actual existance is essentially random. What if a computer could measure the existance or non-existance of something such as that to achieve a wholly random number? Or, perhaps, the decay rate of radioactive elements? I've heard that it is possible to set up an experiment, termed Schrodinger's Cat, wherein a cat is in a sealed box with a gun to its head, and a hair trigger upon the weapon that will trigger if a radioactive element decays. The element has a 50/50 chance of decaying in one hour. The purpose of the experiment is to say that until it is 'observed', at the one hour mark the cat is both alive and dead, and when seen it falls into one state or another. Something to do with quantum realities, perhaps, and that multiple ones of these exist at once, but that we can only observe the probable timeline in which we exist. That is beside the point, however. What I meant by this is that a radioactive element can have a 50/50 chance of decaying. Set up five of these elements within a computer. Say that each of these is a binary number. Thus the ones that decay and do not can set up a binary number that is wholly random to five binal (correct form of the word?) places. Say, for example, 10010. There we have the random number 18. Such a system could give a randomization of up to 31, but a far larger system could be implemented. I am certain there are also other phenomena that exist in the real world that are both measurable and, to our perception and for all purposes, random (it may be that NOTHING is random save for quantum effects, and not even they... but we must look at it realistically. Most of what we see may be causality, but the forces that influence them are so grand, far-reaching, and beyond the influence of Man, that it may as well be random. After all, even shaking and casting lots is essentially dependent on the forces of physics that move about the the pieces, draw one to the ground and make it strike first, etc. A computer, thus, could be essentially random, if not in our day, then in a future world.
Funny thing is, though, can humans do random? If I were to ask someone for a random number, they likely wouldn't go into the millions; if they are in a joking mood, they might, but if not they will automatically bound the answer within reason. If there is something at stake, they will attempt to choose an unlikely number, or a number that has a higher statistically favourable probability. So can humans be random? Do we naturally make use of quantum-type things to come to random conclusions? Essentially, can it not be argued that a human brain must indeed follow a set of rules, or normality, bounded reason, input/output response, etc?
Just a thought. Personally, I'm all for believing in the existance of a soul and the like, and so see things most often in a philisophical rather than scientific light, and am more keen to see beauty than reason, and wondering mystery over full understanding and knowledge.