Deletion (or confirmed re-deletion) of the data is irrelevant at this point, it is the models created from that data, and their use, which will now persist in usefulness to Analytica. Armed with these models, and future refined/iterated versions, they likely will capture the data more directly from users in the future. Once the genie is out, it doesn't readily go back in.
Zuckerberg on Cambridge Analytica situation
161–170 of 583 posts
Re: Zuckerberg on Cambridge Analytica situation
#162Earlier quoted context omitted.
What kind of safeguards would you introduce? At the very start of Facebook platform the API would anonymize the user's email address, the app would get an app-specific hash, e.g. abcdefg-app123456@facebook.net It was a working email address with Facebook handling the forwarding. This proved futile as the very first thing that apps then did was to ask users for their real email address.
To really fix this, Facebook will have to stop allowing 3rd party developers direct access to user data. Basically, FB should introduce an App-Engine like platform where the backend of any 3rd-party application that uses FB data has to run on FB-owned servers. Developers of these applications would then ship their code to FB (similar to Heroku) and run in a sandboxed environment where they are not allowed to take dat…
> the backend of any 3rd-party application that uses FB data has to run on FB-owned servers
Would that work for mobile apps?
Re: Zuckerberg on Cambridge Analytica situation
#163Earlier quoted context omitted.
This was how FB has always designed to work from the beginning Not exactly. In all fairness, as Zuck points out, this is a key part of the story, and in theory, why this is different: In 2015, we learned from journalists at The Guardian that Kogan had shared data from his app with Cambridge Analytica. It is against our policies for developers to share data without people's consent. This raises the question of what Fa…
> It is against our policies for developers to share data without people's consent. Not to mention that it's arguable that "consent" was really given for facebook to share the data in the first place. I'd be interested to see some polling results asking if facebook users knew what facebook was up to and whether they feel OK with it.
I found the old permissions request. Seems fairly clear to me, but I doubt people really understand the consequences
Re: Zuckerberg on Cambridge Analytica situation
#164Earlier quoted context omitted.
What kind of safeguards would you introduce? At the very start of Facebook platform the API would anonymize the user's email address, the app would get an app-specific hash, e.g. abcdefg-app123456@facebook.net It was a working email address with Facebook handling the forwarding. This proved futile as the very first thing that apps then did was to ask users for their real email address.
Then make "no requests for users email" part of the terms of using the API?
Sure, and maybe they'll put it right next to the "you're not allowed to save any data" clause.
Re: Zuckerberg on Cambridge Analytica situation
#165The main problem is that users can't control their data. FB should: - Allow its users to delete their accounts as if they never existed - Show users which person and or service has accessed their data (also FB internally) - Use backend permission management instead of frontend permission management
How about additionally, 1) providing an "Advanced Feed" providing each user a log of every data field exfiltrated, date/time/who/IP addr,/ etc. 2) requiring every single advert and it's targeting parameters, that runs on the system to be made available in a fully searchable advert pool for anyone to search. This would allow people to see what is being targeted, journalists to figure out what targeting is happening (m…
Plus, a lot of the "news feed" articles started as re-shares of microtargetted ads. There absolutely should have been more transparency on virally reshared things that started as microtargetted ads.
Re: Zuckerberg on Cambridge Analytica situation
#166Seems incredibly naive for a company who's core business is data, to grant apps full access to all of friend's data. Did nobody realize that was a bad idea?
I think ambition beat them. The idea was a two-way data exchange; if you use the facebook platform to build apps with user's data,
a. it might be in your interest to help keep the data updated. Say, if your app depends on an accurate home address, you'll ask the user to update it.
b. you might generate more data or content. Say, to publicize your app you might ask the user to post on facebook.
Moreover, it's always advantageous to have other developers depend and build on your platform. It makes your platform richer by adding functionality you may not have the resources to develop, and gives you leverage (ie, Apple and Google's mobile OSes).
If I recall correctly it wasn't "full access to all of friend's data", but similar access to what you had on the user? I thought it was very exciting. I could think of a ton of potential apps, and got to use it at a hackathon once before they limited the access. Back then I thought it was a bummer, but the move made a lot of sense.
Re: Zuckerberg on Cambridge Analytica situation
#167Sorry I'm a fan of FB. I use many of the tools they've provided to the open source world.
I know this won't be a popular opinion but I'd welcome some regulation and bureaucracy for technology firms. This way there wouldn't be this weird gray/imaginary line situation going on that we have right now. I think Congress should probably outsource the regulation and bureaucracy similar to how the SEC works...but maybe not tied to the POTUS and instead creating a structure outside the purview of the POTUS.
Re: Zuckerberg on Cambridge Analytica situation
#168Earlier quoted context omitted.
To really fix this, Facebook will have to stop allowing 3rd party developers direct access to user data. Basically, FB should introduce an App-Engine like platform where the backend of any 3rd-party application that uses FB data has to run on FB-owned servers. Developers of these applications would then ship their code to FB (similar to Heroku) and run in a sandboxed environment where they are not allowed to take dat…
> the backend of any 3rd-party application that uses FB data has to run on FB-owned servers Would that work for mobile apps?
The backend of the mobile app would run on FB servers.
Re: Zuckerberg on Cambridge Analytica situation
#169Earlier quoted context omitted.
> It is against our policies for developers to share data without people's consent. Not to mention that it's arguable that "consent" was really given for facebook to share the data in the first place. I'd be interested to see some polling results asking if facebook users knew what facebook was up to and whether they feel OK with it.
I'd also be really interested to see screenshots of what the users saw when they clicked "ok." I've been able to find a few screenshots online for other apps, but nothing that indicates what it would have looked like in 2013.
https://photos.app.goo.gl/0z3F0BH9Q0wNK0nz1
I think this is from 2013.
Re: Zuckerberg on Cambridge Analytica situation
#170"should not happen again"