4 ms·

I ran a site in the mid 2000s with about 2500 forums. The site barely paid the server cost. I was no Facebook ofc. I had about a million visitors per month. So

I ran a site in the mid 2000s with about 2500 forums. The site barely paid the server cost. I was no Facebook ofc. I had about a million visitors per month. So me not being able to monetize this circumstance lead to me manually fighting spam, which took up all my time. Plus answering support requests and continued development of the site as well as maintenance. I can absolutely understand they automate that. It IS impossible to manage the volume of posts and comments. Even with hired labor Facebook can afford. It's not all that simple as you make it to be. How many moderators do you need to moderate millions of people's posts?

b112 5y ago
Facebook has a much more profitable revenue stream than you ever had, and as I mentioned, the ability to code tools to vastly improve the moderation workflow. There is nothing wrong with automatic flagging, only with automatic action after that flagging. And key here is, a way to legitimately dispute. Your case is not a parallel case.
- Cederfjard 5y ago
  You seem confident in this, so I assume you have run the numbers. How many support agents do you need to moderate 3bn monthly active accounts?
  - bavell 5y ago
    We don't know because they haven't even tried, they'd rather make excuses which conveniently also make them richer.
    - forgotmypw17 5y ago
      I haven't tried it either, but I've tried moderating a decent size community, though nowhere near the numbers of Facebook, of course. There are two challenges which combine into one problem: 1) The amount of bad content posted to a forum rises exponentially with the size of the membership. For every x new people reading, x^y bad content is posted. This is because a large target becomes much more attractive to spammers, trolls, and anyone else looking for a cheap audience. 2) It is extremely difficult to find good moderators. A good moderator will be, within reason, consistent, reliable, AND unbiased. Even with a salary at stake, it is difficult to find people who are even two of those things. In order to succeed in moderating content well, you need to leverage reliable moderators (today, this means rare humans) with unreliable ones (text analysis and less rare humans) through some sort of statistical system. To date, we have not found a reliable system which can work at scale with this problem, but my faith is in a transparent and verifiable web of trust type system. Slashdot does pretty well, but it's still too opaque, in my opinion.
entropyie 5y ago
Similar to arguments about getting human support from Google being impossible... There is a very simple solution: allow users to pay a modest amount, say €25, to get 10 minutes of real human support. This will immediately rule out people spamming at scale. The other side of this equation, is that if the ban is upheld, they must justify it in human terms, beyond : "the ML said it's bad". A positive judgement could then then be used to train the ML to be more lenient on this user going forward, at least when it's a borderline case
- crayboff 5y ago
  That would be immediately seen as Facebook charging people to be unbanned. That would lead to people claiming FB bans accounts to make money.
  - kaibee 5y ago
    > That would lead to people claiming FB bans accounts to make money. More realistically it probably would lead to FB banning accounts to make money. Theoretically you could setup the incentives such that money from banned accounts can only go towards paying moderators though I'm not sure if anyone would believe you.
  - radford-neal 5y ago
    Then have the money go straight to some uncontroversial charity. Facebook still benefits by the reduction in frivolous support requests, perhaps making this a feasible option.
nyuszika7h 5y ago
They don't have to look at every single post manually. But there needs to be a proper appeal process in place for posts/comments flagged as spam by their algorithms that doesn't require making a big fuss about it on social media / HN to get an actual human to respond to you. Big companies like Facebook, Google, etc. would have absolutely no problem affording that, they're simply greedy and prefer to maximize profit at the expense of users.
Beldin 5y ago
Meta just posted billions in profit. Profit is not money spent on improving moderation. With such affluence flowing into coffers instead of investing in sorely needed improvements, GP is absolutely correct in the underlying premise: Facebook/Meta is doing next to nothing about its moderation problem. Even though they can easily afford to put a billion into it.

b112 5y ago

Facebook has a much more profitable revenue stream than you ever had, and as I mentioned, the ability to code tools to vastly improve the moderation workflow. There is nothing wrong with automatic flagging, only with automatic action after that flagging. And key here is, a way to legitimately dispute. Your case is not a parallel case.

Cederfjard 5y ago
You seem confident in this, so I assume you have run the numbers. How many support agents do you need to moderate 3bn monthly active accounts?
- bavell 5y ago
  We don't know because they haven't even tried, they'd rather make excuses which conveniently also make them richer.
  - forgotmypw17 5y ago
    I haven't tried it either, but I've tried moderating a decent size community, though nowhere near the numbers of Facebook, of course. There are two challenges which combine into one problem: 1) The amount of bad content posted to a forum rises exponentially with the size of the membership. For every x new people reading, x^y bad content is posted. This is because a large target becomes much more attractive to spammers, trolls, and anyone else looking for a cheap audience. 2) It is extremely difficult to find good moderators. A good moderator will be, within reason, consistent, reliable, AND unbiased. Even with a salary at stake, it is difficult to find people who are even two of those things. In order to succeed in moderating content well, you need to leverage reliable moderators (today, this means rare humans) with unreliable ones (text analysis and less rare humans) through some sort of statistical system. To date, we have not found a reliable system which can work at scale with this problem, but my faith is in a transparent and verifiable web of trust type system. Slashdot does pretty well, but it's still too opaque, in my opinion.

entropyie 5y ago

Similar to arguments about getting human support from Google being impossible... There is a very simple solution: allow users to pay a modest amount, say €25, to get 10 minutes of real human support. This will immediately rule out people spamming at scale. The other side of this equation, is that if the ban is upheld, they must justify it in human terms, beyond : "the ML said it's bad". A positive judgement could then then be used to train the ML to be more lenient on this user going forward, at least when it's a borderline case

crayboff 5y ago
That would be immediately seen as Facebook charging people to be unbanned. That would lead to people claiming FB bans accounts to make money.
- kaibee 5y ago
  > That would lead to people claiming FB bans accounts to make money. More realistically it probably would lead to FB banning accounts to make money. Theoretically you could setup the incentives such that money from banned accounts can only go towards paying moderators though I'm not sure if anyone would believe you.
- radford-neal 5y ago
  Then have the money go straight to some uncontroversial charity. Facebook still benefits by the reduction in frivolous support requests, perhaps making this a feasible option.

nyuszika7h 5y ago

They don't have to look at every single post manually. But there needs to be a proper appeal process in place for posts/comments flagged as spam by their algorithms that doesn't require making a big fuss about it on social media / HN to get an actual human to respond to you. Big companies like Facebook, Google, etc. would have absolutely no problem affording that, they're simply greedy and prefer to maximize profit at the expense of users.

Beldin 5y ago

Meta just posted billions in profit. Profit is not money spent on improving moderation. With such affluence flowing into coffers instead of investing in sorely needed improvements, GP is absolutely correct in the underlying premise: Facebook/Meta is doing next to nothing about its moderation problem. Even though they can easily afford to put a billion into it.