The scale is exactly the problem. The social contract of what's acceptable for friends, mailmen, or passersby hailing a cab does not apply when extended to automated processes running on computers of global megacorps.
I have no issue with a police officer tailing and tracking a suspect with license plate ABC1234. That license plate number is publicly visible information. I do take issue with the police installing license plate cameras that give them a location history of the entire driving population of a city.
I have no issue with an acquaintance recognizing me by the shape of my face. If I go out in public, I expect to be publicly visible. I do take issue with Facebook running facial recognition on photos taken by their users, or installing their own facial recognition cameras, and sending advertisements and connection suggestions to me based on my proximity to other people.
I have no issue with my neighbor, trying to connect his wifi, seeing my SSID alongside his. The router needs to broadcast a publicly visible management beacon so that he can locate his wireless network and I can locate mine. I do take issue with Google making a global database to allow devices to use the visible SSIDs to determine their location.
I have no issue with a web server querying my browser to see what system fonts I have available, or what size my browser window is. These and other bits of information are useful to display the website I'm requesting to me. I do take issue with the server or their advertising partners using these characteristics to construct a fingerprint to track me across sessions or sites.
I recognize that there's no empirical difference between these uses. Public information is public, it's not obvious why the source should be able to determine how it's used or where the line is between acceptable, intended, private party use and unacceptable, automated, exploitative use. But I think I know what's acceptable when I see it.
Interestingly, the initial example of a street address has always existed in large centralized databases. These databases are even published for mass distribution in the form of a phone book, and they've also been available via online queries for the past 25 years (mapquest, et al)
I have no issue with a police officer tailing and tracking a suspect with license plate ABC1234. That license plate number is publicly visible information. I do take issue with the police installing license plate cameras that give them a location history of the entire driving population of a city.
I have no issue with an acquaintance recognizing me by the shape of my face. If I go out in public, I expect to be publicly visible. I do take issue with Facebook running facial recognition on photos taken by their users, or installing their own facial recognition cameras, and sending advertisements and connection suggestions to me based on my proximity to other people.
I have no issue with my neighbor, trying to connect his wifi, seeing my SSID alongside his. The router needs to broadcast a publicly visible management beacon so that he can locate his wireless network and I can locate mine. I do take issue with Google making a global database to allow devices to use the visible SSIDs to determine their location.
I have no issue with a web server querying my browser to see what system fonts I have available, or what size my browser window is. These and other bits of information are useful to display the website I'm requesting to me. I do take issue with the server or their advertising partners using these characteristics to construct a fingerprint to track me across sessions or sites.
I recognize that there's no empirical difference between these uses. Public information is public, it's not obvious why the source should be able to determine how it's used or where the line is between acceptable, intended, private party use and unacceptable, automated, exploitative use. But I think I know what's acceptable when I see it.