Dear Alexey,
I see, thanks. You say: For instance, if all the world with all our knowledge about it were only a dream or a computer simulation of a joker from the upper level, this "knowledge" would not value much, would it?
Let me call this position (V). I'd actually challenge (V). If all the knowledge obtained by observation and thought were not knowledge about the world but about a dream or a computer simulation, I see two possibilities:
(1) If it were in principle possible to find out, by means of observation and thought, that our knowledge is not about the actual world but just about a dream or a computer simulation, then we could eventually reach this point and start to peer behind the curtain ('escape from the matrix').
(2) If it were not possible at all to find out, by means of observation and thought, that our knowledge is not about the actual world but about a dream or a computer simulation, then so what. We would simply continue to increase our knowledge about this dream or computer simulation. Actually in this case, we might as well call that dream or computer simulation "world" instead, because that would just be a matter of nomenclature.
You employ an argument similar to (V) in the next step, when you conclude that thought (and also goals?) can't emerge from some lower level, mechanical or aleatory, don't you? I'm asking, since in my essay I do actually explain how goals might emerge at macroscopic scales.
Cheers, Stefan