Hacker News Reader

Comment by proksoup | Hacker News Reader

> The details of moderation practices are routinely hidden from public view, siloed within companies and treated as trade secrets when it comes to users and the public. Despite persistent calls from civil society advocates for transparency, social media companies do not publish details of their internal content moderation guidelines; no major platform has made such guidelines public.

It's really nice to see light shining on this topic. Truly, this content moderation is the equivalent (equivalent for the internet medium) of the boundaries of debate enforced by the editors and owners of major media corporations on our previous mediums.

View on HN

If the details of moderation rules were public, companies would be facing additional labor costs dealing with posters pushing hard on the edges of those rules.

At some level, users can choose from a variety of forums with different moderation methods, or none, and companies can try to fit their moderation policies and practices to their users' expectations.

All of that develops in a cross between evolution and arms race, with lots of moving targets - global variation, changing social mores, changing technology, the latest taboos and moral panics.

Exposing the details of how the sausage is made would help the people who want to put pebbles (or worse) in the sausage more than it would help the makers or consumers. If one finds the sausage too bland, there is likely a spicier one just a few clicks away. Or one could grind one's own sausage and compete!

What you describe sounds like security by obscurity.

Publishing moderation policy details would be like crowd-sourcing cryptography software development - creating a fractally increased attack surface with a less profitable business model.

Social media companies operate in a market economy. These are not public utilities or government programs. Feel free to propose a more cost-effective approach.

I think exposing the rules will make them stronger.

Moderation is not cyber security. Virii don't turn around and argue that your AV is unfairly applying their rules to them.

I think the rules are quite different when dealing with human processes rather than algorithmic ones.

Could you elaborate, the same principle seems to apply to me, the costs of dealing with people attacking the edge strengthens the edge, no?

Or plausible deniability.

If the details of moderation rules were public, companies would be facing additional labor costs dealing with posters pushing hard on the edges of those rules.

If the rules have really hard edges. I think that's very unlikely. I think the substantial reason companies won't publish their moderation rules is almost certainly that they don't have much more than a few bullet points - and a page of "things the lawyers have said you mustn't let people say" - and otherwise leave it up to the whim of the moderators.

Having said that, it might well be illegal - or in contempt of court - to publish the list of things that are moderated for legal reasons.

Reply on news.ycombinator.com