Salem woman denied use of Facebook because of Native American name

5 points by dbg31415 ↗ HN
At what point do we just call Facebook racist? They really don't seem to be taking any action to fix discriminatory policies.

* Salem woman denied use of Facebook because of Native American name - KPTV - FOX 12 || http://www.kptv.com/story/28853494/salem-woman-denied-use-of-facebook-because-of-native-american-name

* Facebook's Name Policy Strikes Again, This Time at Native Americans | Electronic Frontier Foundation || https://www.eff.org/deeplinks/2015/02/facebooks-name-policy-strikes-again-time-native-americans

* Petition · To Allow Native Americans to use their Native names on their profiles. · Change.org || https://www.change.org/p/facebook-to-allow-native-americans-to-use-their-native-names-on-their-profiles

8 comments

[ 4.3 ms ] story [ 30.4 ms ] thread
You may call Facebook racist right now, if you would like. But let's look at this from another side shall we? Facebook has approximately 9000 employees. They have approximately 1.4 Billion users. Or in other words, they have about 1 employee for every 156,000 users. This means, they don't actually keep track of every user that shows up to their site. They couldn't. They would have to have every employee in the company look at a user every 10 seconds, every moment of every day to look over the users in a one year time frame.

So, what they actually do is automate the system and attempt to catch people who are not following their rules using algorithms. A human never reviews such a case at all. If someone is blocked it's because some piece of code decided they are trying to game the system. It doesn't just catch Native American's. It probably catches thousands or more accounts per day of people who actually are trying to game their system. What you are reading about is a handful of false positives on an otherwise successful system.

We also need to remember that users like these are NOT Facebooks customers. Facebook attempts to be nice and fix the problem and actually it seems like they make things right more often then not. But these users are Facebooks product, not their customer. Their customers are advertisers who want to put their advertisements in front of people. They treat people really well for not really needing them at all.

Using algorithms to limit the amount of work facebook employee's have to do to make sure users are following the rules makes perfect sense. It doesn't explain why you would cut humans out of the process entirely.

To your point about users are not Facebook customers is technically correct, but these users are actually the product that facebook is trying to sell. If you were a company looking to buy ad space would you use a platform that mistreated your potential customers?

Both good points. I happen to think their algorithms are flawed, by the way, but that may be beside the point. And I think they should ponder having an appeals process, but again that is at their discretion. They are not a public service or a government agency, they are a private organization with actual concerns about the legitimacy of their user base. I don't think they have the people-power to handle the number of requests they would get in a day. How would you suggest they change that?

I entirely agree that the users of Facebook are their product, and really in this case they are doing quality control. A manufacturer of widgets must check the widgets on the way out of the factory door on the way to the consumer to make sure that the widgets are in good working condition. If they are not, they must refund the consumer, or they will lose their reputation as a widget producer and potentially their entire business.

The same goes for Facebook. They are selling eye balls and they want to tell their customers, the advertisers, that the eye balls that they are buying are actual people and not robots, scripts, or repeats of other eye balls (people with multiple user accounts under different names_. These algorithms are a quality control mechanism, nothing more nothing less.

As for the question of mistreating potential customers. That is a complex question. I would say there are probably numerous companies that do not advertise with Facebook because of how they treat users. On the other hand, there are obviously enough advertisers who don't care at all how Facebook treats their users and will throw piles of money their way to show their ads. Ultimately I would imagine that most advertisers who do spend money with Facebook want two things. First they want real, vetted users. They would probably prefer that Facebook screw up some minor tiny fraction of their users base in the hopes of cleaning up the exceptional majority of "ad reading" users. Second, advertisers want people who will open their ads. Click through rates on Facebook are abysmal as it is. Having real, vetted users in theory should raise they rate, not lower it.

Oh, and with all of that, I have to say no, they are not racist, sexist, or any other -ist. They honestly don't care who you are. If you fit the profile of someone that their advertisers are willing to pay them to put ads in front of, you will be allowed into the system. It is purely a matter of economics in a capitalistic society.
Yeah because the economics of a capitalistic society are never racist...

And, more to the point, if Facebook designs an algorithm for validating names that assumes that names not conforming to particular cultural norms are not valid, that is in fact racist.

I am sure there are times where economics in a capitalistic society is racist. But I do not believe that is the case in this instance with Facebook. No one has yet proven that they are discriminating against a group of people. That would require that you come up with information which told us how many Native American's are actually being discriminated against. The articles name maybe a dozen or so people effected by this policy (which is rather low I am sure there are more than that, to be fair). Let's assume that all of these people were Native American and do some back of the envelope math to see what percentage that effects.

World population = 8,000,000,000 Native American population (2010 Census) = 5,000,000 Facebook users = 1,400,000,000 Percentage of world users who are Facebook users = 17.5%

If an even number of people across all nationalities use Facebook this would mean there are: 875000 Native American Facebook users, give or take.

That means that: 0.0014% of Native Americans are troubled by this policy according to those who have stepped forward in these articles.

Of course there are probably many many more. 10 times more? That would be 0.014% of all Native Americans. 100 times more? You get the idea.

On your second point you state that if Facebook designed an algorithm... are you sure they checked every possible cultural naming norm when they designed the algorithm? I do not believe they did. Does this make them discriminatory or racist? In fact, no. The real question here is, do the update their algorithm to become smarter and better at allowing in real users and keeping out fake ones? The answer here is yes. That would make them the opposite of racist. In order to be discriminatory, you must actually discriminate.

Facebook does discriminate. Look at what they do to small businesses that do not pay for ads and only want organic (read: free) advertising. They are very open and public about the fact that they do not care about and do discriminate against these small business individuals.

It's more like calling an English spellchecking algorithm (or its authors) racist because it doesn't recognize some words. It could be racist, but it could also just be that writing the software is difficult. And validating/recognizing names is a well-known AI problem.

I'd rather just call all of this anti-human, because it's actually about empowering robots and monetizing humans.

(comment deleted)