Dear Hrvoje Nikolic,
at first glance your contribution to the contest maybe for some readers seems to be a successfull exercise in "laziness" of typing some more words into the machine. But contemplating your essay for a while, it has a certain economical elegance and deepness for me that at least i can't deny. So at least for me there are good reasons for your essay to have been considered as eligible in this contest, for initiating a deeper thinking about what could be possible/impossible and what could be the borderline between these two seemingly mututal exclusive alternatives.
So let me contribute some additional remarks on your assumptions.
I am sure that you are aware of the problem of the frequencies in QM if extrapolated to yet unseen universes, subsets, events etc. So you are right to state that some of them could be very unlikely. But that surely does not contradict yours statement that anything could be possible. This statement is presumably based on the framework of QM *and* human experience of the fact of (more or less) surprisingly changes in the world's evolution (including maybe human evolution, scientific intelligence and consciousness in general).
Theoretically, there could be possibly one exception of your claim, namely if it could be also possible that at some day we detect an *objective* way (personally i don't believe that this will happen and it is hard to imagine for me, but i can't exclude this possibility) to show that one of our theories (Many Worlds, String theory etc.) is indeed the final one (ToE). It is hard for me to imagine that this sort of scenario will happen, cause my conviction is that for such a scenario one must transcendent all reality in one's own consciousness to the utmost borders to come to the final information how the complete landscape of reality looks like. Well, as said before, i can't exclude a scenario in which - for whatever reasons - humanity comes to a point at which all people can suddenly transcendent the whole of all_that_exists. In my opinion this could - logically - only be possible if all_that_exists would consist only and only out of the stuff that transcendents it's own existence (namely consciousness).
But even in the latter case a problem remains open: The question of wether in this consciousness there is a fixed set of all "thinkable" things (that would mean that there has to be at least one "thinkable" thing in this set that would contain the whole set within itself [infinitely?]) for the purpose of being able to grasp it and would for the sake of my argument incorporated at some point of the thinker's evolution into the thinkers consciousness - or that there is a continuum in the property of "real" that can evolve to more or less "real" things. For the first case, it is hard for me to imagine how a conscious entity can draw an objective and definite border between the things that are "thinkable" and the things that aren't "thinkable", because that would mean that for drawing that border at the right and "objective" place, this entity had to think not only of all "thinkable" things but also of all "unthinkable" things to draw the border correctly. The latter would be a striking contradiction to the term of "thinkable".
But that wouldn't really mean that this couldn't be possible, it could *probably* only mean that *I* - in my anthropic view - can't imagine it!!!
So for the latter case and also for the case where a conscious entity can't draw a border between the thinkable and the unthinkable, there could be a certain empty space and room for creative change, development and self-multiplicity for the one thing we are searching for: reality. Could it be that "reality" is a creative thing, creative in such a manner that in it's whole potential there could be true *all* things, even the impossible ones - and even the contrary and (QM) probabilities are a reflection of the whole system's (in the sense of all_that_exists) framework of *values* at some point of time?
I guess, if these lines of reasonings would be true, our meaning of the word "true" would become a more relative one instead of an absolute one (the latter we commonly conclude a priori in science). At least it would fit into human experience of all of our subjective truths and different value systems. But maybe the latter is also only anthropic reasoning extrapolated to the whole cosmos and if this would be the case we would be back at the point we started from.
My conclusion is, that for me it is at least remarkable that natural systems bear some kind of self-referential loops and anyway (or even therefore?) are stable enough to produce repeatable, predictable behaviour of matter and at least some (half-)truths. And within "matter" i would include also the human brain.
