[imgabe]

This seems very lazy on Facebook's part. Rather than employing people who will research and fact-check, they're punting the problem to the users. This whole situation arose because users are very bad at determining which sources are credible. People will believe anything that confirms their biases, and disbelieve anything that doesn't.

[gervase]

I think this is spot on, but it misses the core reasoning: this approach permits them to look like they're doing something, without actually changing the status quo. In other words, this lets them have their cake and eat it too, because users (and news orgs) can't blame Facebook for the rankings - they're "from the users"!

At least, that's how this news reads from my admittedly-cynical perspective.

[Reedx]

How would that work with people who think the New York Times is fake news to begin with? They're not going to suddenly change their mind because fb says they're real. It'll simply reinforce people's mind or they'll say fb is biased or paid off or whatever. Even if fb had a perfect system it wouldn't solve the problem. People didn't use logic when coming to the idea that NYT is fake and Breitbart is real, so you can't use logic to get them out of it.

It's a no win situation at this point. I'm not sure there's any solution short of removing news from the platform.

[matt4077]

I'm sure Facebook wouldn't care about a few million spent on a panel of the world's greatest connoisseurs of real news.

This feature is in fact guaranteed to cost far more than a few experts ever would.

But Facebook knows that any expert panel will without a sliver of doubt quickly converge to a ranking that has the Economist and the New York Times in top positions, and Breitbart somewhere behind a random word generator.

Any working statistical method will lead to the same result, obviously. But it gives Facebook the option to invoke HN's favourite argument: "it can't be political because it's the algorithm. See: here are numbers."[0].

[0]: Compare, for example, the libertarian love for bitcoin, and how it's free from the politics that undermine Central Banks,

[peoplewindow]

Any working statistical method will lead to the same result? What do you mean by that?

I read articles in the New York Times, the Economist and also Breitbart (UK edition) fairly regularly. Whilst they cover very different stories as you'd expect given their political biases, I have not noticed any major difference in accuracy when dealing with objective facts. This is partly because "mainstream" media is quite unreliable, rather than any awesome quality of reliability inherent to Breitbart, but the idea that it's unreliable seems to be to be coming from people who simply dislike conservative worldviews ... and desperately want to stop people from reading them. Same reason they try and smear anyone who goes looking as nazis, bigots etc. They fear that if someone reads things from the "other side" they might find it's not so unreasonable after all ...

[ben_jones]

It's also a huge amount of hubris. You either punt this problem to the users in which case you have massive, incurable, vote brigadding, or you pretend to be opening it up "democratically" but in reality you use heavy handed algorithms to curb all outliers in whatever bias the algorithm deems fit.

[swarnie_]

Its also a lot cheaper then hiring people.

[ghostbrainalpha]

It's not the cost of hiring people they really need to avoid.

They don't want to alienate the conservative crowd who basically check in to Facebook ONLY to see the day's new anti-Hillary meme.

I don't know about you guys, but my Grandma clicks on ads at an almost bot like rate.

[nerfhammer]

because then people will accuse the fact checkers of being biased and carrying out agendas and conspiracies.

if there's no particular overlord that people can point to then they can't be as outraged at it.