Enola KnezevicMy browser, and pretty much everybody's browser, has a unique digital fingerprint. Why on Earth do the same websites then need to repeatedly check that I am not a robot??? If I wasn't a robot yesterday, I am not one today either!
Bitan Nath@rhelune This is an interesting problem! Most companies (linkedin etc) utilize deeper versions of https://github.com/fingerprintjs/fingerprintjs (a larger bit hash). IG your solution is to have an IDFA like large bit hash that refreshes per time unit+website for zero collisions (everyone is unique and human, privacy is provided by the OEM, nobody can track across websites, or assimilate behavior across hours). Only an OEM that doesnt make money from data collection has incentive to.
@bitanath Not sure I understand and don't have the time to look through the library. I kind of expect every website to have their own cache of confirmed humans. Or they could even use non-refusable essential cookies.
@bitanath The thing is, they use that unique identifier, but only for evil (advertising, surveillance), not to make the user experience better.
@rhelune yeah my tl:dr; 1) Reply: They cannot build a reliable cache without login (anonymous browsing+gdpr)
[Assuming objective is never being tracked while anonymous]
2) My scenario 1: Either everyone is unique and cannot be tracked because the OEM keeps changing their uniqueness (no CAPTCHA ever)
3) My scenario 2: Everyone is the same therefore cannot be tracked (always CAPTCHA) since bots and humans are also the same
Either is valid but it’s an interesting problem!
@bitanath My understanding was that they create our pseudonymised shadow profiles anyway.
@rhelune Ah! No.. they could, but it wont be reliable. You can uninstall an extension, install a new font, latch onto a different cell tower, restart your home wifi… etc.. Their math depends on larger and larger bit hashes, to reduce the impact of the randomness of your actions. To break that, either make every bit of your hash random, or make no bit hash of anyone random lmao 
@rhelune The alternative is to normalize all features across all browsers and extensions. Hide in the crowd (no one is unique, no one is guaranteed to be human, no one is private). However, there is no point in tracking right now since everyone looks exactly the same. In such a situation, only a CAPTCHA is proof of humanity (sure this can also be simulated etc, but the problems can also get harder). While this can be used to identify human, all humans look the same.