What would also be interesting is training real cephalopod neural networks in the lab on various type of synthetic input and see what kinds of data they can lea

What would also be interesting is training real cephalopod neural networks in the lab on various type of synthetic input and see what kinds of data they can learn to make sense of. Need better brain-computer interfaces I guess - perhaps something like neural lace. Cephalopod powered skynet, here we come.