> But the study revealed that a third of the connections in the fruit fly’s brain did not follow this pattern—they were between two axons, between two dendrites

> But the study revealed that a third of the connections in the fruit fly’s brain did not follow this pattern—they were between two axons, between two dendrites or from a dendrite to an axon. Interesting. Does this extend to humans? Does it offer a plausible biological mechanism for backprop?