3 ms·

I'm just a software engineer. I'm sorry you think that I'm doing a poor job of defending Facebook, because I think in that a lot of the criticisms are inaccurat

I'm just a software engineer. I'm sorry you think that I'm doing a poor job of defending Facebook, because I think in that a lot of the criticisms are inaccurate and come from a poor understanding of the privacy policies, the motivations and implications. Admittedly, this is mostly because they are insanely complicated, and we need to do a much better job educating people. We (where that includes me and all the people I know at Facebook) care very much about doing the right thing for users while still innovating and giving users the most utility we can. I know it's easier to just think of us as an evil corporation out to suck your private secrets and sell them to the slimiest buyer around, but that just isn't true. [Where there are valid criticisms, you can be sure we're taking them very seriously.]

martey 16y ago
While the motivations of the creators of the privacy policies and options might be pure, the fact that the settings are confusing enough that people are exposing information that they want to keep private is pretty damnable. If Facebook cares about protecting the data of their users, making the privacy settings more intuitive would be more productive than worrying about needing to "educate" users. This includes having default options that hide data instead of making it public.
- nfnaaron 16y ago
  "the fact that the settings are confusing enough that people are exposing information that they want to keep private is pretty damnable." The information is being exposed, not because of confusion, but because Facebook wants the information exposed (to business partners and affiliates). The confusion leads to surprise, not exposure. "making the privacy settings more intuitive would be more productive than worrying about needing to "educate" users." This was exactly my thought earlier in the thread. If it's this hard to understand and explain, there's some simplification yet to be done.
- indigoviolet 16y ago
  Of course it would be best if the settings were so simple that no education were necessary. But my claim is (a) for a site like Facebook to work "well" [where we can argue ad infinitum what "well" means], some information needs to be shared in the system, and some needs to be private. (b) this is often different for different groups of people. You now have to implement a privacy system that works for every piece of content, for every set of users that I want to define a setting for. I think you'll quickly come to something fairly complicated. There are very smart people trying to come up with abstractions and interfaces that make this easy to do, but I fear that at the end of the day, users are simply going to have to learn some new tricks if they don't want to be surprised. And that means, we have to figure out how to explain this complex system to them.
iamdave 16y ago
You brought up an interesting point. come from a poor understanding of the privacy policies, the motivations and the implications What are the motivations? The key, primary, and focal motivations behind all of these changes?
- indigoviolet 16y ago
  Frankly I'm a little hesitant to attempt to explain in too much detail, because I don't want to do a bad job and I'm not sure I'm seeing all the angles. I'll try to do a very high-level view. Keep in mind that this is just my perspective. The most important thing is that although we started as a very simple profile-plus-friends social network, there are now a very large number of people connected in this manner. We have the opportunity for the first time to really let a user define their identity in one place [their relationships to real people and entities in the real world, their likes and dislikes], and shape the content on the web around that identity. We have the ability to share that opportunity with smart hackers everywhere -- but we obviously cannot just willy-nilly hand over the data to anyone. We need to define a usable API, a set of acceptable practices, a way to enforce these things so we respect the user's trust without completely destroying the app developer's ability to do something useful. This is really hard. We're slowly finding our way to doing this well, but there are many unforeseen issues, especially because app developers have a lot of money at stake. We're iterating on our API, trying to change up the constraints on developers. At the same time, several cool applications wouldn't be possible if everyone were to keep everything private from everyone -- you have to be social to some extent on a social network if you are to derive any value from it. So we're trying to get people using this complicated privacy interface where you can control publicness/privateness at a very fine grained level, but you need to understand many things to do this correctly. We're also trying to set things up so that you can hide things as you want from the real people who might judge you for it or whatever, but at the same time, you can get functionality related to that interest from some application [for example, you may not want to tell your friends you like Britney Spears, but Pandora should still be able to play you Britney Spears and similar music]. There's a whole other dimension related to the Open Graph stuff -- building a "semantic web", having one graph where users and entities are connected and organized in a way that can be leveraged to do all these cool things that the web originally promised but hasn't been able to deliver. There are tremendous engineering challenges to putting all this together. As you can see, this is a lot of balls to be juggling. Our philosophy has always been to iterate quickly, correct missteps just as quickly, and not be afraid to innovate. There are a few things I'm pretty confident we don't ever want to do-- the first of which is that we don't want to ever sell user data. That would be an incredibly stupid business decision for a company that wants to help you manage your identity on the web, and we recognize better than anybody that users need to trust us. I think I've gone on long enough-- I'm not sure I actually answered your specific question. Hopefully my blathering was useful to give you some feel for our motivations though. If you want to learn more, you can listen to some of the talks here: http://apps.facebook.com/feightlive/ http://apps.facebook.com/feightlive/ . I recommend the closing remarks by Chris Cox.
jfager 16y ago
Here are my criticisms, would you please let me know which ones are inaccurate? If you don't want to explain fully why each one is inaccurate, you can just give me the numbers, and I'll do my own further research. 1. New features are rolled out without advance notice and without prior explanation of their privacy implications. 2. The privacy policy changes without advance notice, often towards ensuring less privacy. 3. Features that expose more personal information are frequently enabled by default (opt-out rather than opt-in). 4. With instant personalization, social plugins, and the 'likes' feature, FB has switched to making no distinction between 'public if you look for it' and 'publicly broadcast'. 5. These features also indicate that FB has started tracking users browsing habits across the web. It's not enough to know everything about a person's social life, FB wants to know as much as it can about their private life, too. 6. FB asks to inspect user's email address books while suing a company that uses the same basic technique to manage FB profiles on the user's behalf. This is staggeringly hypocritical. 7. That same Friend Finder feature does not explain what gets pulled from your email account, and it does not adequately explain that those email addresses are stored and may be used not only to help that user connect with people who already have FB accounts. It does not explain that FB treats the information it pulls from email accounts as a shadow social graph. 8. There is no easy way to export user data. 9. The privacy settings are lengthy and confusing, and there's no simple way to restrict everything in one go. To ensure your privacy, you literally have to wade through page after page, and you have to do so on a relatively frequent basis as new features can undo your work (see #1,2,3).

martey 16y ago

While the motivations of the creators of the privacy policies and options might be pure, the fact that the settings are confusing enough that people are exposing information that they want to keep private is pretty damnable. If Facebook cares about protecting the data of their users, making the privacy settings more intuitive would be more productive than worrying about needing to "educate" users. This includes having default options that hide data instead of making it public.

nfnaaron 16y ago
"the fact that the settings are confusing enough that people are exposing information that they want to keep private is pretty damnable." The information is being exposed, not because of confusion, but because Facebook wants the information exposed (to business partners and affiliates). The confusion leads to surprise, not exposure. "making the privacy settings more intuitive would be more productive than worrying about needing to "educate" users." This was exactly my thought earlier in the thread. If it's this hard to understand and explain, there's some simplification yet to be done.
indigoviolet 16y ago
Of course it would be best if the settings were so simple that no education were necessary. But my claim is (a) for a site like Facebook to work "well" [where we can argue ad infinitum what "well" means], some information needs to be shared in the system, and some needs to be private. (b) this is often different for different groups of people. You now have to implement a privacy system that works for every piece of content, for every set of users that I want to define a setting for. I think you'll quickly come to something fairly complicated. There are very smart people trying to come up with abstractions and interfaces that make this easy to do, but I fear that at the end of the day, users are simply going to have to learn some new tricks if they don't want to be surprised. And that means, we have to figure out how to explain this complex system to them.

iamdave 16y ago

You brought up an interesting point. come from a poor understanding of the privacy policies, the motivations and the implications What are the motivations? The key, primary, and focal motivations behind all of these changes?

indigoviolet 16y ago
Frankly I'm a little hesitant to attempt to explain in too much detail, because I don't want to do a bad job and I'm not sure I'm seeing all the angles. I'll try to do a very high-level view. Keep in mind that this is just my perspective. The most important thing is that although we started as a very simple profile-plus-friends social network, there are now a very large number of people connected in this manner. We have the opportunity for the first time to really let a user define their identity in one place [their relationships to real people and entities in the real world, their likes and dislikes], and shape the content on the web around that identity. We have the ability to share that opportunity with smart hackers everywhere -- but we obviously cannot just willy-nilly hand over the data to anyone. We need to define a usable API, a set of acceptable practices, a way to enforce these things so we respect the user's trust without completely destroying the app developer's ability to do something useful. This is really hard. We're slowly finding our way to doing this well, but there are many unforeseen issues, especially because app developers have a lot of money at stake. We're iterating on our API, trying to change up the constraints on developers. At the same time, several cool applications wouldn't be possible if everyone were to keep everything private from everyone -- you have to be social to some extent on a social network if you are to derive any value from it. So we're trying to get people using this complicated privacy interface where you can control publicness/privateness at a very fine grained level, but you need to understand many things to do this correctly. We're also trying to set things up so that you can hide things as you want from the real people who might judge you for it or whatever, but at the same time, you can get functionality related to that interest from some application [for example, you may not want to tell your friends you like Britney Spears, but Pandora should still be able to play you Britney Spears and similar music]. There's a whole other dimension related to the Open Graph stuff -- building a "semantic web", having one graph where users and entities are connected and organized in a way that can be leveraged to do all these cool things that the web originally promised but hasn't been able to deliver. There are tremendous engineering challenges to putting all this together. As you can see, this is a lot of balls to be juggling. Our philosophy has always been to iterate quickly, correct missteps just as quickly, and not be afraid to innovate. There are a few things I'm pretty confident we don't ever want to do-- the first of which is that we don't want to ever sell user data. That would be an incredibly stupid business decision for a company that wants to help you manage your identity on the web, and we recognize better than anybody that users need to trust us. I think I've gone on long enough-- I'm not sure I actually answered your specific question. Hopefully my blathering was useful to give you some feel for our motivations though. If you want to learn more, you can listen to some of the talks here: http://apps.facebook.com/feightlive/ http://apps.facebook.com/feightlive/ . I recommend the closing remarks by Chris Cox.

jfager 16y ago

Here are my criticisms, would you please let me know which ones are inaccurate? If you don't want to explain fully why each one is inaccurate, you can just give me the numbers, and I'll do my own further research. 1. New features are rolled out without advance notice and without prior explanation of their privacy implications. 2. The privacy policy changes without advance notice, often towards ensuring less privacy. 3. Features that expose more personal information are frequently enabled by default (opt-out rather than opt-in). 4. With instant personalization, social plugins, and the 'likes' feature, FB has switched to making no distinction between 'public if you look for it' and 'publicly broadcast'. 5. These features also indicate that FB has started tracking users browsing habits across the web. It's not enough to know everything about a person's social life, FB wants to know as much as it can about their private life, too. 6. FB asks to inspect user's email address books while suing a company that uses the same basic technique to manage FB profiles on the user's behalf. This is staggeringly hypocritical. 7. That same Friend Finder feature does not explain what gets pulled from your email account, and it does not adequately explain that those email addresses are stored and may be used not only to help that user connect with people who already have FB accounts. It does not explain that FB treats the information it pulls from email accounts as a shadow social graph. 8. There is no easy way to export user data. 9. The privacy settings are lengthy and confusing, and there's no simple way to restrict everything in one go. To ensure your privacy, you literally have to wade through page after page, and you have to do so on a relatively frequent basis as new features can undo your work (see #1,2,3).