Facebook is rating the trustworthiness of its users on a scale from zero to one

Post by cabalamat » Tue, Aug 21, 2018, 5:48 PM UTC

mipmap04 wrote: post #33

vorpalhex wrote: post #16

> Because Facebook forwards posts that are marked as false to third-party fact-checkers, she said it was important to build systems to assess whether the posts were likely to be false to make efficient use of fact-checkers’ time. This seems like a "the road to hell is paved with good intentions" kind of thing. On one hand, I understand not wanting to forward every single story when a troll is mass-reporting, on the o…

What would be a viable alternative? I don't think they can just ignore the issue and, as you mention, can't review every single story. They'll have to rely on some form of sampling for the reviewers. Would it be better to just take a truly random sample of all reports? I don't think that would be very effective given the number of publications they have to review. Do you let users review the content for veracity? See…

Post by austincheney » Tue, Aug 21, 2018, 5:51 PM UTC

igravious wrote: post #83

danielblazevski wrote: post #43

Earlier quoted context omitted.

Good intentions? I would even argue against that. I find it hard to believe facebook has good intentions with this. Remember, Mr. Zuckerberg is alleged to have called Mr. Trump the night of the election to congratulate him. Let’s stop pretending like the tech industry is full of good intentioned people making an “oops”

As a European I want to make it vocally known that in my opinion US politics has in recent times almost irrevocably poisoned HackerNews; and that is coming from someone who has been on this site eight years. One _cannot_ in _any way_ infer a lack of good intentions from a single alleged phone call. Besides, if I were a titan of industry I would do my best to call whoever the hell got elected president the night of th…

Post by degenerate » Tue, Aug 21, 2018, 5:54 PM UTC

panarky wrote: post #91

ChuckMcM wrote: post #59

I am tempted to paraphrase the Elephant man, "I am not a sigmoid function, I am a human being!" in response to that headline :-) That said, I think it is great that they are trying to characterize this sort of thing. The second law of institutions is that every process must require more people to execute than can conspire without being detected[1]. But relying so much on humans in one way or another is going to gum i…

> every process must require more people to execute than can conspire without being detected Step 1 - let users report posts that are false Step 2 - human fact checkers investigate user reports Step 3 - rely more on users who report posts as false that really are false, and less on users who report posts as false that really are true Step 4 - repeat step 3 but for the fact checkers themselves Step 5 - train an ML mod…

Post by a235 » Tue, Aug 21, 2018, 5:57 PM UTC

reaperducer wrote: post #32

vorpalhex wrote: post #16

> Because Facebook forwards posts that are marked as false to third-party fact-checkers, she said it was important to build systems to assess whether the posts were likely to be false to make efficient use of fact-checkers’ time. This seems like a "the road to hell is paved with good intentions" kind of thing. On one hand, I understand not wanting to forward every single story when a troll is mass-reporting, on the o…

third-party fact-checkers This is the part that stood out to me. It's allows Facebook to tell lawmakers and the public "Hey, look! We're taking things seriously!" Until something goes wrong. Then they can lay the blame on someone else and issue another non-apology apology saying, "We can do better."

Post by cheeseomlit » Tue, Aug 21, 2018, 5:57 PM UTC

luka-birsa wrote: post #98

Is it anything like the Chinese credit system? Seems the two could share the notes.