Related? Directed complexes in the brain have huge redundancy (compared to random graphs) possibly to permit uncorrupted signal transfer <a href="https://www.birs.ca/workshops/2017/17w5043/files/Pawel_Dlotko_directed_complexes.pdf" rel="nofollow">https://www.birs.ca/workshops/2017/17w5043/files/Pawel_Dlotk...</a>
This could lead to extremely interest and important conclusions. I now see two variants.<p>Variant 1: all our neuro-computing industry just on wrong way (may be this is intentional trick by hardware manufacturers, but scientists at field could also missed something important in pursue for low hanging fruits), and we already missed point where consciousness happen, and instead we unnecessary spent huge resources on communications on language understandable for our computers (for benchmarks understandable for investors).<p>Variant 2: large Neural Networks already consciousness, but we don't understand them, and they looks on us like playing Dolphins (or cats, if you love cats). And once we will be extremely discouraged, when they will just ask something like "human, you really want to continue play this game of words? How about make something serious?"