Maybe:
* This account already scored poorly on spamminess and attempting to post a bare link (with no content) on their page pushed them over the edge?
* The EFF Privacy Tips link was somehow used in a parallel spam campaign, or some characteristic of the page itself causes it to be flagged as spammy?
* Something about this specific browser session was flagged, preventing posting? (this might go along with the No Violations observation about the page).
I find it much more interesting to try to reverse engineer these strange behaviors of some spam detection systems than to just chalk them up to a "Facebook hates the EFF" conspiracy theory - if this happened to me, I'd be doing some more tests to try to see what's up.
Some part of me wonders how is this even news? I guess "some person was blocked temporarily and that's all we really know" is not going to make it to the top of HN.
this is far too little data to make assumptions like the one in the link.
But alas, we humans love our conspiracy theories because they tell a more interesting story.
Exactly! This is also one of the most traditional domain names (Creation Date: 1990-10-10T04:00:00Z). I can assume FB has a lot of people working on spam prevention and they should have reviewed all old domains, relevant NGOs, political parties, country-level governments, etc. by now and chose to not whitelist EFF. If it was malice or incompetence it is a matter of (endless?) discussion, however.
You know, ultimately, you can choose to not share that much with Facebook. There are tools for privacy that don't let Facebook get too much from you. The easiest one is clearing cookies automatically in a browser to disrupt long-term tracking. Beyond that, there are VPNs and burner phone numbers. I feel like people are not as powerless against Facebook as you might mean to say, and it's not that hard to care about privacy and value Facebook.
Facebook has a very long history with privacy that speaks for itself.
I disagree with your sentiment. Being public/private doesn't make one more/less trustworthy. I would posit it is the business model of the company itself that should be given scrutiny.
The point here is that facebook is particularly abhorrent. It's a risk to offload any data to anyone outside of your own self, but it's a serious mistake to interact with a zuck product in any way.
No, this is just an observation with no strong emotions or reactions attached - it's tiring and it's probably unhealthy. No one has wronged me or caused offence to me. If anything, it's self-inflicted. We all ultimately decide what we consume.
Anyways, your comment was interesting in the way you see this, and I think, illustrative of the point I was making. Outrage online has become so normal it is to be expected.
By that I don't mean it is bad, just that it is business as usual for the EFF. The article is expected and not particularly provocative. There are dozens like this on the EFF own Facebook page.
That's what makes me think the content of the article and the fact it comes from the EFF doesn't have much to do with the blocking.
https://gist.github.com/smashah/667d4d5cf31670ee87547450861a...
Actually it's not at all funny, it's horrifying
It's getting a bit depressing that the only thing worth replying to so many comments is 'Bot.'
And without more evidence, that seems more plausible to me than Facebook banning links to the EFF.
That's 10 false positives _every single day_. And in all likelihood their spam detector isn't that good, and there are many, many more posts than that.
Every now and then one of those false positives will be an interesting web site or one that feels really obviously wrong. But that's just what statistics does. When you take enough samples from a distribution, even very low-probability events happen.
Sometimes something will feel really out of the ordinary and wrong, but it happened entirely by chance.
If there were evidence of systematic decisions like this, it would be more of a story. But what we have here is just a big nothing-burger.
The reality is that moderation relies heavily on imperfect machine learning models and overworked human reviewers making rushed judgments on hundreds of cases per day. There's no meticulous strategy document mapping out the pros and cons before banning accounts that upset the company.
Mistakes inevitably happen when relying on this combination of flawed automation and human reviewers who are stretched too thin. The moderation policies may seem arbitrary or politically motivated from the outside, but much of it comes down to hasty human error and buggy algorithms rather than some malicious scheme.
well it's subtly wrong but points in the overall direction.
the capacity to manipulate this kind of stuff, to sabotage ideologies such as EFFs privacy ideology IS the product.
governments spend money for this, it's a product from governments to other governments.
though I suppose this is an instance of FB dogfooding
