
A “perfect interface” would be one that effectively taps into the incredible capabilities people have for processing natural scenes. For example trees are incredibly complex, yet their structure is easily discernable.
I’m not saying that we should slavishly copy nature and make every interface look like a landscape. But if we look at the specific mechanisms that humans use to extract information from the scenes we were evolved to understand, we may be able to tune our interface techniques to show us things in the way we can grasp them most easily.
For instance, fruits ripen along a green-yellow-red spectrum–spanning the most recently acquired color channel (green to red) that our eyes use to encode color. We could use this deeply-wired continuum in an inventory system to show maturing, ready-to-sell stock, as red. This would make it stand out, as well as unconsciously saying “ripe & ready to pick.”
Brad Paley

My work in making information more meaningful to people revolves around two key ideas: that information should be presented in a way better tuned to human perceptual abilities (for example, I typically make “information objects” rather than tables). And the best way to make information more easily understood is to design the representation around the way that the user thinks.
This image contrasts a standard table, which must be read to be interpreted, with a table of information objects. The shapes and colors are tuned to the specific job: blue=buy, red=sell; left=buy, right-pointing=sell; dark=new, light=seen; no color=completed. In this way we get the perceptual system instantly to do things that perviously had to be extracted by reading. The most important function demands the most visual attention, assuring attention where it’s needed.
Brad Paley