Hacker News Reader

Comment by dofm | Hacker News Reader

The chosen thresholds are a bit absurd — e.g. if it was five rather than two I think authentic users would barely notice — but this actually might be what is needed to make Facebook useful again for people.

The amount of fake link spam, "link in comments" spam, misleading shit, is endless.

No amount of anti-spam technology is going to help. Cold hard cash requirement might do something about it.

If you assume that the only reason to run a fake celebrity page, for example, is to mislead users into some action, this might help there. Though actually what is needed there is a similar price pattern on posting email addresses.

View on HN

Don’t forget the other big fraud vector:

Bot 1: “I have been using the services of Mr Bob John who is a financial advisor, he has made me $70,000 in the past six months. Very trustworthy”

Bot 2: “Oh wow! Will you please share how I can contact him?”

Bot 1: “sure, you can message him on Telegram. His username is BobJohn983739257”

Bot 3: “I have also used Mr Bob John services. He has been a reliable and faithful advisor.”

Yeah — this is largely the dominant evolution of 419 fraud isn't it? Rarely very bespoke scams.

I think you could more or less instaban users who mention Telegram and nothing of value would be lost.

I see comment likes and replies from users called "Congratulations you've won…" or "Message me to win your prize" and if you report those, they are actually pretty much always removed (easy pattern I guess).

But nobody seems to have any luck with reporting fake celebrity scams, to the point that if you are a well-known user who has more comments than you can reply to, you more or less have to regularly post a video saying "I will never message you directly".

> I think you could more or less instaban users who mention Telegram and nothing of value would be lost.

Oh for sure. But they'd just switch to saying "†e|_e gr@mm" or something. It does seem to me like the cheapest models today should be cheap and good enough to finally kill off this type of spam, though, if deployed correctly.

My preference would be that spam comments detected in this way get a bit set on them that causes them to be visible only to their posters, and posting more than one just causes all your future comments to also get that bit regardless of their content, but since it's a shadowban, it's harder for spammers to know that their account is burned.

> The chosen thresholds are a bit absurd

The threshold isn't for your grandpa, it's for Facebook Pages, which are basically landing pages for businesses on Facebook. And I think it's chosen very deliberately: so that you can set up your page for free, promote it, and then realize you need to pay to keep it alive.

> The threshold isn't for your grandpa, it's for Facebook Pages

Yes, I understood this, thanks.

It doesn't at all change my point: one of Facebook's biggest problems is deceptive pages that clone user content for clout, mislead users and then ultimately direct them for scams.

If you set a limit of five posts a month without paying, you would perhaps only punish fake pages, which are a high volume game.

And my point is that they don't want to punish scammers, they want to get money from businesses. In that view, the limit is not absurd, it's pretty carefully calibrated to make sure that every business with an active Facebook Page needs to pay (but not before they get set up).

Reply on news.ycombinator.com