Re:Facebook collects a wide range of private data from devel...
by edmundsauto on Friday February 22, 2019 @07:38PM
(#19230643)
A Bayesian analysis would have to include all the decisions where FB was a good steward of user data. The events reported in the news are a very small fraction of the possible times that FB could have done wrong.
Not that I'm defending FB, but your attempt to lend credence to your statement with a smart sounding approach was undercut by selecting a superficial and biased prior.
Facebook's claim about not wanting the data is contradicted by actions. They chose to make their SDK send[1] analytics signals on library init before the user could have even been presented with a request for consent. They chose to have their analytics SDK send[2] everything to Facebook by default, requiring developers to go out of their way to disable the spyware (including somehow discovering that this step is needed).
> Now if Facebook were to use this data for their own purposes,
What would Bayesian analysis say about that question given a history with multiple events where FB et al were using the all of data they received however they want? Facebook lost the benefit of the doubt a long time ago, and it will take a lot of work to rebuild their reputation.
[1] https://media.ccc.de/v/35c3-9941-how_facebook_tracks_you_on_...
[2] Ibid.