I re-downloaded Google Earth recently. Currently Google would like you to view Google Earth as a Chrome web app. You can get around this by downloading the 'pro' version, which is a desktop app on Windows, or you can possibly download the desktop app in your Linux package manager, if you use that operating system (it is present in Linux Mint's software manager). Google have been 3D-mapping large parts of the globe lately. You might have seen this by activating the satellite layer in Google maps if you view the desktop version in a browser (the feature does not appear to be present in the mobile app, or at least not the versions I have seen). Not everywhere has been mapped yet, so it might not cover your local area. Within Google Earth, if you zoom right down to the surface it will take you to Street View if you zoom in over or near a road and that road has Street View data. However, if that area is off the road, or does not have Street View data, then it will put you into a mode called 'ground-level view'. You can also switch from Street View to ground-level view by pressing a building icon, next to the man icon, in the upper right corner of the window. Ground-level view is the feature that I want to talk about. If you are in ground-level view with the 3D-buildings option selected in the left margin and the area has been 3D-mapped, then you can see a 3D-model of the area. This is apparently derived from satellite LiDAR data, which creates a fairly crude 3D-model, and satellite photographs are imposed over this model as surface textures. What makes this creepy is that you can nose around people's properties. You can take a virtual walk around someone's back garden, no matter if it is walled-off or not. It does not cover my area, but my father's house is included, and I could look at his house from any angle. It is a bit crude, as mentioned, but it is nevertheless representative of the house. I could look around his neighbours' houses as well. I saw, for example, that one had a swimming pool with a slide and a wagon wheel hung against the wall of the building in his walled garden (the house is located abroad in a Mediterranean country). I wonder if this feature will last. I am surprised there has not been any publicized backlash against it. There were widespread reports back when Google Maps first came out, with it allowing people to get satellite imagery of the grounds of people's houses, and again when Street View was released, with it allowing people to look at the anyone's house from the street. The ground-level view feature has existed within the app for a while. I can find mentions of it in blog and forum posts as far back as 2010. However, the prevalence of large 3D-mapped areas was not as great back then, if it existed at all. It seems in the old days the ground-level view was restricted to showing coarse topographical data and user-supplied 3D models of notable buildings. I have not heard anything about ground-level view in the press. Ground-level view is essentially the privacy infringement of Google Maps' satellite layer and Street View combined. Street View enables people only to see the street-facing side of most people's houses, which is publicly visible anyway. But ground-level view enables people to walk around a virtual model of someone else's property, even if those parts are normally walled-off from the public. To get an idea of the current quality, here is the British Library in London: I did not want to single out someone's private residence for the sake of their privacy. One wonders what this will look like in a few years' time. Google will probably be able to look through people's windows. The talk we used to hear about twenty or more years ago about satellites being able to read newsprint from orbit greatly exaggerates their capabilities, especially at that time. Sharper detail is however possible with multiple images of the same subject if it remains static for a long period. Those features which do not move can theoretically be imaged at resolutions much higher than the satellite's normal resolution by using pictures from successive passes to construct a more detailed image using a method known as frame stacking. The resolution of the 3D model can be refined with more LiDAR data and better LiDAR sensors on future satellites.