Facebook's Fix-It Team

http://fortune.com/longform/facebook-fix-it-team-fortune-500/

by sethbannon • 8 years ago

46 38 8 years ago

Today I learned: "anyone convicted of two or more murders is banned from Facebook." (One murder is okay though.)

yial3• 8 years ago

I think most interestingly is "... (The reason: While people may commit a single homicide accidentally or in self-defense,..." homicide would be unlawful killing, so now I'm curious, is this per incident... as in, imagine someone had to kill two individuals at once in self defense, would they be banned from facebook?

sidebar: I have had my facebook deleted since 09/2015 so I may not be the key audience ever of facebook. Also, I am not condoning deadly force used legally or illegally, I am just curious as more of a thought experiment.

gnarbarian2• 8 years ago

Key word there is convicted. Homocide in certain situations can be legal. Killing someone in self defense is legal (in the us). Another example of homicide with no fault to you would be someone jumping out in front of your car. Neither case would normally yield a conviction or even a case.

Also homicide is different than murder. Murder implies intent.

JshWright• 8 years ago

Homicide is not necessarily unlawful. It simply means someone was killed by another Homo sapiens.

There are all sorts of situations where that could occur lawfully.

cowpig• 8 years ago

warning: auto-play video

duxup• 8 years ago

"In the spring of 2017, Rosen put a team of engineers on “lockdown,” a Facebook practice in which people drop everything to solve a problem. The problem was dire indeed: People were using Facebook Live, a video-streaming service that had just launched, to announce their intention to kill themselves, and even to stream themselves doing it."

While I've got some strong feelings about Facebook and their actions... that's got to be a frightful experience, creating a system where you realize people are sharing their suicide on. There's nothing about being a technology company that makes you capable of dealing with such an issue but there it is and now you have to act.

sanderjd• 8 years ago

This is the kind of empathy that I think lots of people fail to have when it comes to sites with user-generated content. As another example, lots of people seem to think Facebook is negligent for failing to properly delineate between fake and real news articles or between organic political speech and propaganda. Purely from an engineering perspective, I just don't see how they could possibly do even an adequate, let alone a good job of that.

borplk• 8 years ago

It's only recently that people have started to assign blame and responsibility to the platform hosting the content (really only after the US election drama).

Back in the earlier days of the internet the attitude was quite different.

It was all celebrated free speech and "anyone is empowered to do anything it's all open and free and happy for everybody and there's some bad stuff so what" (happy early Web 2.0 days where the web was moving from static pages to "textboxes and submit buttons").

Now a pin drops and 50 publications write an "Facebook negligently allowed someone to ... they must be held accountable and regulated and broken apart" article.

It's just a mindless pendulum swing. Those of us in the middle always get the shit. Back in the day we got blamed for saying anything against the "open and free and rainbows for all humanity" rhetoric and now it's gone to the other side.

deanCommie• 8 years ago

Facebook used to have a feature where when you wrote a comment to a news article, it would give you suggestions of 3 "suggested" response links that you could post in response.

99% of the time when I started writing a response to a news story I knew to be false, the top recommended link was a Snopes article debunking the story. I could just click on that, and be done with it.

So they already had an engineering fix for this problem, and it worked GREAT.

Why did they take it out? Probably because the metrics showed that stuff like this killed conversations, and the virality of people reposting stories. Nobody is going to Share a news post when one of the top comments is a big Snopes "FALSE" preview.

Some engagement metrics went down and they took it away.

pentae2• 8 years ago

Well from an engineering perspective you could create a blacklist and a whitelist. There's not a huge amount of variance in news sources out there, i'd say 80% of news sources being shared on social media are less than 100 domains. The rest of the fake news sites domains could be added to the blacklist based on a combination of user reporting and moderation. You could literally put a team of about 5 people on this and over a long enough timeline it would be solved.

gkanai• 8 years ago

creating a system where you realize people are sharing their suicide on.

They should have planned for that pre-launch and had a plan in place to deal with users like that. When Live was launched, journalists asked Zuckerberg about this exact topic and he would not believe the Live platform would be used in that way. Of course murders and suicides were streamed on the platform and FB was yet again on the back foot.

They continue to make 'mistakes' in understanding human nature. FB plans for the optimistic usage when in reality, that is rarely the normal usage.

aylmao2• 8 years ago

FB plans for the optimistic usage when in reality, that is rarely the normal usage.

Hold up, this sounds wrong. I am sure the optimistic usage is by far the "normal usage"; I don't suppose most people are streaming murders, suicides, etc.

Considering the number of users Facebook has, I'd be surprised if these negative uses even come close to one in many thousands. For the same reason one also can't assume they won't happen; give 2+ billion people the ability to stream live and someone is bound to stream something nasty. I doubt most of those 2+ billion people are using this feature negatively though.

chrisseaton• 8 years ago

They should have planned for that pre-launch

That's easy to say with hindsight isn't it?

exikyut• 8 years ago

When Live was launched, journalists asked Zuckerberg about this exact topic and he would not believe the Live platform would be used in that way.

The other comments in this thread have discussed the "optimistic usage". Well, it's not really politically correct to say "yes, that's likely to happen, and we've implemented a bunch of fixes to deal with that", regardless of whether it had happened.

Deploying Hanlon's Razor, I'm thinking it was foreseen but not taken as seriously as it should have been.

The other possibility is that, it was foreseen, the reaction was expected, and the burning-towers-of-money-to-fix-all-the-things response is marketing at the end of the day to make the site seem responsibl{,y scalabl}e.

fyfy18• 8 years ago

Is there a way to pass the GDPR consent dialogue without giving them permission to do whatever they like with my data? I thought the purpose of the GDPR was to give users a choice as to whether their data is used?

scaryclam• 8 years ago

Some may be in breach of the GDPR for making consent a condition of use. Without knowing more it's hard to say though, since some services really do need some PII to operate (like Amazon needing an address to deliver to).

bertil• 8 years ago

Facebook’s GDPR modal explains what they do with your data and why it’s legitimate. You can opt out of targeted advertising. The only thing that Facebook doesn’t offer is not filtering your News Feed by interest, which they see as a key feature.

makomk2• 8 years ago

Facebook's GDPR modal may do, the one on the linked Fortune article simply doesn't allow you to access Fortune unless you agree to let them use all your personal information for advertising.

billysielu• 8 years ago

When I click the link to this article all I see is a terms and conditions document which I don't agree to. I should not have to sacrifice privacy just to read a public web page.

The content pasted below for those who can't see it / already agreed it:

About Your Privacy on this Site

Welcome! To bring you the best content on our sites and applications, Meredith partners with third party advertisers to serve digital ads, including personalized digital ads. Those advertisers use tracking technologies to collect information about your activity on our sites and applications and across the Internet and your other apps and devices.

You always have the choice to experience our sites without personalized advertising based on your web browsing activity by visiting the DAA’s Consumer Choice page, the NAI's website, and/or the EU online choices page, from each of your browsers or devices. To avoid personalized advertising based on your mobile app activity, you can install the DAA’s AppChoices app here. You can find much more information about your privacy choices in our privacy policy.   Even if you choose not to have your activity tracked by third parties for advertising services, you will still see non-personalized ads on our site.

By clicking continue below and using our sites or applications, you agree that we and our third party advertisers can:

    transfer your personal data to the United States or other countries, and
    process your personal data to serve you with personalized ads, subject to your choices as described above and in our privacy policy.

thecatspaw• 8 years ago

To avoid personalized advertising based on your mobile app activity, you can install the DAA’s AppChoices app here.

this is ridiculous. Is this even GDPR compliant?

krageon• 8 years ago

No, because this is opt-out behaviour. It needs to be opt-in: To avoid personalised advertising, your actions required should be "none". It should work like that by default.

peterjmag• 8 years ago

Agreed, that's obnoxious. I bypassed it with Google's cache: http://webcache.googleusercontent.com/search?q=cache%3Ahttp%...

dpark• 8 years ago

I agree that this is pretty obnoxious, but this is also clearly not a “public web page.” If access is not open, I don’t see how you could call this public.

AndrewKemendo• 8 years ago

For 20 minutes on the morning of May 1, Facebook users saw a curious query at the end of every update on their feeds. “Does this post contain hate speech?” they were asked, in small font next to “yes” and “no” buttons.

I once believed this ultra basic, explicit supervised reinforcement learning query system would be the future of user interfaces and AI. That if you asked people to close the loop in a learning system, some number of them would give good feedback and you could make an increasingly better decision loop.

What early tests like these seem to prove, and maybe I'm being pessimistic, is that people generally don't want to help make systems better (with "better" defined loosely). Even if they do want to help, users either can't differentiate their own perception or the interface doesn't give clear enough directions.

It seems to boil down to users and the companies not being explicitly value aligned. Or rather, when they need to take an action which reveals an explicit value proposition, the user responds to the emotional/signaling request with noise, rather than data. As a result the users don't take it seriously and so they break the system. It seems to be built into humans to reject the concept of providing explicit value input (aka voting on if something is good or not) into a system that you haven't been totally indoctrinated into - like voting for politicians. I would also guess that it being tied to your person makes that even less likely to happen.

cjhopman• 8 years ago

Yeah, that sounds totally logically sound. It explains why yelp reviews don't in any way reflect the quality of businesses and why aggregated movie ratings don't reflect the quality of movies.

Facebook's Fix-It Team

Related Stories