Other friends have had the same problem with actual child pornography.
But... that's precisely what the article is about?
> The Korean agent showing Janczewski around eventually pointed the phone’s camera at a desktop computer on the floor of Son’s bedroom, a cheap-looking tower-style PC with its case open on one side. The computer’s guts revealed the hard drives that Son seemed to have added, one by one, as each drive had filled up with terabytes of child exploitation videos.
> This was the Welcome to Video server.
> "I was expecting some kind of glowing, ominous thing," Janczewski remembers, "and it was just this dumpy computer. It was just so strange. This dumpy computer, that had caused so much havoc around the world, was sitting on this kid’s floor."
The incentive in this case was profit from/access to more CSAM. GP is referring to how most subreddits/forums do not have that incentive.
It provides a larger target where failure of the system to regulate content can impact more users.
But it also provides economies of scale in the regulation apparatus having outsized impact when it is improved.
The tricky question is whether the mute grows proportional to the megaphone.
> This all worked fine pre-internet. Editors would decide who and what would be posted.
Obviously this argument is inconsistent.
It's also wrong. The worst problem isn't with Reddit, it's with sites like the one reported in the article. These are exactly like the "under censored ones [sites]" you advocate for above. They don't lose users - they gain the types of users who want that kind of material.
What if there were no redit servers, and each subredit was self hosted? What would the incentive be for people to post child porn on some small independently run server?
As the subredit owner, you could just censor whoever you wanted, with very little global internet drama. Tools for autoflagging posts could still exist. Over-censored forums would piss off users, and be easily replaced. Similarly, under censored ones would lose users / get busted.
This all worked fine pre-internet. Editors would decide who and what would be posted.
It still works for things like audio fiction podcasts, where "slush readers" filter submissions, and money / reputation are exchanged when stuff gets published. In addition to leading to much better moderation than the big Internet sites, it creates jobs for domain experts.
The main problem is that it doesn't scale to a $1B/mo business that investors can siphon money off of. Instead, it's best structured as an ecosystem of small botique businesses.