[aryoung1]

We discuss it at the end of this blog (https://eyes.mit.edu/what-if/), but the main future goal would be to discover new types of eyes or vision systems. The results we show on the website generally agree with the conclusions from biologists, but what if we put the animal in an environment not actually feasible in the real world? Like what if we put an animal on a mars-like world, what kind of eyes would it evolve? And if we constrain that animal to only evolve eyes that we can manufacture (like a camera), would it discover a new type of camera or algorithm that's better than what humans have engineered? So here we show an attempt to recreate biological vision, but we're interested in applying it to artificial vision (i.e., computer vision, robotics, etc.).

[1970-01-01]

Maybe I just can't see the big picture (pun++) but why limit the simulations to the eye? Many other senses are simultaneously working for survival. I'm not aware of anything that just uses the eye for survival. https://en.wikipedia.org/wiki/Sense

[aryoung1]

That's very true and is also an area of interest. The number of possible animals that can be evolved increases substantially with more parameters, so adding more sensors/types of sensors makes the problem even more difficult. Also our animals are simple balls, where most intelligent animals have arms, legs, so on. So modeling more would definitely be interesting, too!