OP's point is that given that the brain does it, and given that it is possible in principle to simulate the brain (even if we don't know how yet), the result of

OP's point is that given that the brain does it, and given that it is possible in principle to simulate the brain (even if we don't know how yet), the result of such a simulation would necessarily be "AGI", disproving the original claim that "AGI is mathematically impossible".