Drawing inspiration from early cartographers who had to make maps with limited information, Outside Text tested models on world map output, also with limited information.
In the earliest renditions of the world, you can see the world not as it is, but as it was to one person in particular. They’re each delightfully egocentric, with the cartographer’s home most often marking the Exact Center Of The Known World. But as you stray further from known routes, details fade, and precise contours give way to educated guesses at the boundaries of the creator’s knowledge. It’s really an intimate thing.
If there’s one type of mind I most desperately want that view into, it’s that of an AI. So, it’s in this spirit that I ask: what does the Earth look like to a large language model?
Prompting “draw a world map” would have yielded obvious results, so to test, a grid was entered, and the probability of land in each cell was calculated.
