In fairness, there’s a lot of technology behind GeoSynth, which will feed a huge range of geotagged photos into a Virtual Earth and PhotoSynth mashup.

Seems the idea is that if enough people take a picture of a particular location, PhotoSynth can build up a pretty realistic digitised version. Stitch thousands of these together and you have something a bit like Street View, except without Microsoft driving around taking photos itself.

“The system would take the best images from a location to create a single image of a specific landmark,” said a Virtual Earth tech specialist.

Sounds pretty cool if you like the whole concept of street-level views of locations. You’d probably get some locations (cities, famous landmarks) with a huge amount of photo data while others would be fairly limited, but then Google’s only really mapping large cities at the street level.

Being able to run a virtual tour of various landmarks just from photographs and without the need for any tags (apart from the geolocation data) is a great idea as this PhotoSynth video demonstrates.