Are you asking if random actions can lead to an action in a certain
direction then the anser is yes. For instance Brownian motion under a
chemical reaction at far equilibrium can posess a directionality.
Prigogine and others have looked at processes at far equilibrium and found
that they can lead to increase in order and complexity. Furthermore under
natural selection, random mutations can get selected according to
survivability and certain mutations will propagate and others will not.
JQ: I get your reasoning that bad mutations would not propogate, but the
chances become even more 'out of this world' each generation that some
random mutation will actually improve upon what was previously improved
upon.
How come that mutations become harder as each generation moves on ?
JQ This type of lottery type improvement would had to have happened
for millions of generations.
But it is not lottery improvement but selection of favourable mutations
over time. And indeed it requires a lot of time and generations. That is
why Darwin assumed that several hundred million years was a minimal
requirement for evolution.