Bing Maps: New York as Canvas
translation from lab to large teams making a giant product
what does it take to make many many people create something interesting that allows for creativity inside and outside?
so: microsoft's success was based on enabling an ecosystem of developers
team size: when things get large, they slow down -- ie, communication dominates doing --> heat death
mapping online
- first oriented to search and directions -- but that's just one application
-- all the interesting ideas end up below the cutline, and you never get to them
image-based rendering for ground level imagery (which is much more information-intensive, changes more frequently, has more interesting applications)
- blurring difference between map and virtual reality
showing Bing -- trying to pick image sources -- aerial photos, synthetically generated 3D images, street side images -- dynamically based on what kinds of information make sense based on scale
-- handling transitions between image tiles and image sources
-- impressive "swooping down" transition from aerial to street side
- debate about building models vs "capturing a city" -- this is image-based rendering, so based on photos stretched over 3D geometry used for semantics and transitions. "In a standstill, you are looking at real imagery."
- fusion of two different sorts of capture -- topdown, structured, corporate and unstructured, from users -- so structured capture is where you really have to worry about privacy
-- structured capture is important in urban areas as a "trellis" for applications
-- so faces and licensed plates are blurred in structured capture images "believe me, I've had many discussions with lawyers about this"
integrating photosynth with the map
- photosynth as user-generated virtual environment
- trying to fuse the user-generated photosynth with the map and making the "transitions as cinematic as possible"
enabling map apps - such as hyperlocal blogs
Q: What skills do you need to make a map layer?
Q: What about mobile phones?
Q: Natalie - conversation about symmetric access to information about you
A: Think about how Amazon handles things -- by seeing the inputs to the record and letting people remove things from it "let me see what the surveillance camera thought of my butt" [when trying jeans on] -- but this all requires engineering commitments.
