This isn’t a “data breach.” And the underlying data gathered isn’t that scary. It was completely legal to collect under Facebook API rules through 2014 (2015 if you count the sunset period)
Show this thread
Some perspective on this Guardian/NYT Cambridge Analytica Story.http://www.nytimes.com/2018/03/17/us/politics/cambridge-analytica-trump-campaign.html …
-
-
-
Numerous campaign Facebook apps in the 2012-14 period collected data about Facebook friends. The Obama campaign called it targeted sharing.
Show this thread -
Where CA screwed up was in transferring the data from an academic. This was a clear no-no under Facebook’s platform policies. Had they simply collected the data themselves, they wouldn’t have had a problem.
Show this thread -
In theory, these apps were not supposed to make use of friends data globally (only as part of the individual friend’s UX) but numerous apps did show aggregate data about friends of friends.
Show this thread -
In fact, the Obama campaign ran and publicly reported analytics about “friends of friends” through its apps — reporting that only 50% of millennials’ friends were contactable via phone but something like 90% were via Facebook
Show this thread -
Off a base off 1M authorized accounts, they estimated they could contact 98% of the US Facebook population via friends. That’s only knowable if you have the IDs in your database.
Show this thread -
What was creepy and scary when Trump did it (actually no, because they didn’t use the data) was heralded as an amazing tech innovation when Obama did it — which it was
Show this thread -
Facebook shut this down in 2014, citing user feedback. More likely, they realized it was possible for third party developers to build massive microtargeting databases this way, competing with their business model.
Show this thread -
And for good reason. Partisanship models based on Facebook page like data are as reliable (if not more reliable) than voter files.
Show this thread -
But you don’t really need Facebook data to replace voter file data (which is good). You need it to replace third party commercial data (which isn’t that good, because of privacy worries) = $$$
Show this thread -
Needless to say, this was problematic for Facebook.
Show this thread -
Needless to say, you shouldn’t be particularly concerned about this. Because the psychographic use case is and was pretty weak sauce. Ultimately you need offline commercial databases to make predictions about the entire population...
Show this thread -
And these offline variables don’t have any predictive power when predicting personality traits — which in and of themselves are not even that useful to know for political persuasion
Show this thread -
To recap: This wasn’t that controversial when it was happening. Facebook shut it down in 2014, partly for privacy reasons, but more for commercial reasons. It’s only scary ex post facto with the words “Trump” and “Russia” applied.
Show this thread
End of conversation
New conversation -
-
-
You're making a number of leaps here. 1. To definitively say they didn't use the data when people associated with the campaign have bragged about the effectiveness of microtargeting *and* the Special Counsel is looking into this exact question is beyond overstatement.
-
2. To dismiss the idea that any of this would work at all anyway is pretty disingenuous when CA's entire use-case is based on the theory that *it does* & they've spent the last 2 years flying around the world selling it.
-
3. CA has 2 pretty good case studies they've been using to sell it in Trump2016 &
#Brexit. 4. You're essentially arguing that emotional advertising doesn't work. 5. You're acting as if data sophistication didn't evolve exponentially between Obama 2008 & Trump 2016. -
So based on your Tweets in sum, either CA is lying their way around the world about their own efficacy & selling vaporware to governments & militaries or there's some absolutely enormous pieces you're missing here.
-
Agreed. Someone should ask CA what they are up to in MEX
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.