Why don't octopuses have a blind spot like humans do?

Because their eye is wired the other way round. In vertebrates like us, the optic nerve has to pass through the light-sensing photoreceptor layer to leave the eye, which leaves a natural blind spot. In cephalopods such as octopuses, the retinal axons pass over the back of the retina instead, so the optic nerve exits without ever crossing the photoreceptor layer, and there is no blind spot at all.

That's strange, because a cephalopod's eye looks a lot like a human eye at first glance: an iris, a round lens, a gel-filled chamber, and light-sensing cells feeding signals to the brain through an optic nerve. For 140 years, scientists have pointed to this resemblance as a textbook case of convergent evolution, two very different lineages independently arriving at the same basic design. But the resemblance only goes skin-deep. A cephalopod's eye develops by folding inward from the body's surface rather than growing outward from the brain, so its cornea just sits on top rather than being structurally part of the eye. Instead of changing the shape of its lens to focus, the way a human eye does, a cephalopod eye physically moves its spherical, fully internal lens back and forth, more like adjusting a camera or telescope. Even the lens's basic building blocks evolved independently: the proteins involved appear to have arisen separately in cephalopods and vertebrates, not inherited from a shared source.

Cephalopod eyes carry their own extra tricks, too. Octopuses have an automatic response that keeps their pupils always horizontal. Squid and octopuses, and possibly cuttlefish, can even detect the orientation of polarized light, something human eyes simply can't do, likely useful for spotting prey, navigating, or signaling to other color-changing cephalopods. Because the cephalopod eye is so analogous to ours, researchers now use it to study gene expression in the eye and eye disease, a more accurate model than fruit flies, and because cephalopods can regenerate their eyes, the same animal can be studied across several trials.