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

That would be immediately seen as Facebook charging people to be unbanned. That would lead to people claiming FB bans accounts to make money.

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.

> 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.

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.

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.

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?

We don't know because they haven't even tried, they'd rather make excuses which conveniently also make them richer.

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.

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.