nothing against a narrow scope, but i find it interesting that there's not even a comparison in accuracy/efficacy in the information provided to the user between starting with the baseline AI and starting with the "wayfinding" AI (or at all). maybe i missed that part? is this essentially just a UX study?
