So imagine you type "cat". Then maybe you get a picture of a cat. Maybe from ancient Egypt. Or maybe tiles of labeled behavior videos, sleep, walk, eat, etc. Or a yowl. Maybe with a scrubbable intensity plot. Or maybe that intensity is shown as a fast spreading sound ripple, overlaid on a top view of your own neighborhood. Maybe with a draggable microphone, to play with volume, and delay, and doppler. Or back to cat, maybe you get video with a 200 bpm cat's fast heart rate. And maybe an order of magnitude animal mass slider. Slide up to human, to whale, and get the word "whale", video of swimming whale, and a whale's 20 bpm heart swishing. And similarly down to mouse 600 bpm. Or back to cat, ... it's a high-dimensional densely-woven tapestry of possibilities.
When the most popular introductory astronomy textbooks can't even manage to get the color of the Sun right, there seems a need to explore new ways to deliver an understanding of the physical world. Pity child automatic speech recognition is still so poor.
Back down towards MVP, might "type a word, get a picture" be interesting? Not so much? Way too distracting? Other thoughts? Thanks for input.