An interesting thread popped in my Facebook feed today. AI pundit Stephen Little shared a ChatGPT Image 2.0 travel map prompted by DAR member Katherine Borges.
It caught my attention (despite having incorrectly plotted data) because the map was gorgeous and the infographics were so effective.
Everything just works. The minimalist terrain map with a compass and concise legend / key insets. The elegance of the notched corner double outlines echoed from the whole map to the insets; complimented by antique styled horizontal rules. The subtlety of the lightened knockouts for all the overlays and terain-deflected travel-lines.
I would be thrilled to embed such an infographic in a Gramps generated Genealogy book. It looked like something a professional illustrator would have crafted for a history textbook.
Stephen suggested that generating a KML to feed into an online map might have been a better choice (at this evolutionary stage) rather than trusting a halucinatory AI. However, a reply pointed out that generating and feeding KMLs to mapping services is not yet a feat for mere mortals.
Perhaps Gramps could bridge these barriers? Gramps can already generate a KLM (with accurate plotting data) and perhaps Gramps could also feed it directly to a mapping service for an interactive map.
Or as part of an Gramps composed AI prompt that outputs a repeatable style of illustration. (Gramps doesn’t simplify Place hierarchies very well yet. But the AI could be prompted to do that that part too.)
