To add to this great thread, here is the Lenski experiment (with the accompaning creationist fiasco), and how it's going so far. Here is another article on how it's going so far.
I just want to thank cofty for his patience on teaching this subject. It is difficult and unbelievable to grasp at first, but if you are really thirsty for knowledge, you will keep searching and find the answers.
As for the typewriter or computer analogy, perhaps there is a better illustration. I don't intend to say one, because of the limitations of analogies. All you have to remember is the accumulation of mutations that are sustained, sometimes because of fitness, sometimes random, always because of natural selection. The monkeys are typing away without any purpose in mind. The typewriter and the paper don't have a purpose either. But imagine this weird scenario. For the sake of comparing it to biological systems, imagine the paper gives 'birth' to a typewriter, and that the mutation that the paper first gained is now passed along to the typewriter, so that a key originally had is replaced with the 'mutation'. This 'mutation' is but a replica of another key. As an example, imagine your very own keyboard with which you write to communicate on this forum. The next keyboard that is replicated has a mutation, and your 'r' key is now a 'w'. Your keyboard now has 2 w's, and the monkey is now more likely to hit that w that will help complete the phrase you are looking for (the outside, non-participant observer). While you are looking on the outside, even though you look for a desired outcome, the process is random, and repeats itself multiple times. The outside observer in this case is the scientist, observing the experiment and looking to see how such complexity could have come about. The monkeys type away, the keyboards press upon the papers, the papers then 'spawn' the mutated keyboards which the monkeys type away at, again and again, some accumulating the mutation you, the non-participant observer, are looking for. Some generations of paper accumulate the mutations you are looking for, others do not. In fact, some harmful mutations prevent the keyboards from typing at all, and so all the accumulation of mutations for that particular keyboard end with that keyboard.
Tying it in with the Lenski experiment, which I highly recommend you read about, as it introduces novel complexity through beneficial mutation accumulation, and once and for all trumps the 'irreducible complexity' argument, the keyboards finally are able to produce the sentence "Methinks it is like a weasel" on a paper. Many typewriters came and went, and so did many papers, but at the end, this particular 'strain' of keyboard/paper accumulated what was once thought irreducibly complex. They key (pun intended) is that each successive generation successful of reproduction passed on this beneficial mutation, and then the other one added another beneficial mutation, culminating in the final product.
I could go on, but see how the analogy start to break down after a certain point? Even so, just because the analogies fail to grasp the total reality, they get us a little bit closer to understanding this seemingly complex process, to the point where you can see that, indeed, the accumulation of mutations. The important thing to understand is this is real life. Lenski mapped this process out from beggining to end. The e. coli bacteria had no way of metabolizing citrate before these mutations accumulated in its DNA. Then, all the right 'ingredients' came to together in one generation and, wham! The bacteria population that was now able to feed on citrate, on top of on meager doses of glucose, blew up to enormous proportions, compared to other strains only feeding on glucose. This bacteria had a selective advantage in now have two sources of energy with which it could continue to survive on and reproduce. These genes were passed on, and here is an experiment showing evolution before our very eyes, much to the chagrin of stubborn creationists.
Mutations don't immediately produce Boeing 747's, but they eventually produce all the novel complexity and diversity we see in the natural world.