Have they decoded whatever the first layer is coming straight out the eye? It would be interesting to know if its roughly the same across different groups of species; do all insects share a similar retinal output, do all mammals?
Anyway, once that initial decode is possible— which implies we can encode, too—hook I/O into lab mice or flys or whatever and hack your way into the next layer (whatever it is that biologically decodes retinal output, and the signals produced), and then hopefully continue to apply this method until you hit a point where you have to start decoding other brain subsystems because its interacting with memory and body position info and etc.
