[protomyth]

You are correct, publishers don't have a legal requirement to be unbiased, but publishers can be sued. Service Providers cannot be sued for what appears on their service[1]. The argument is that "biased" decisions on what to publish makes them publishers and not service providers.

1) ok, in the US anyone can get sued and there seem to be some specific things people can sue and win on, but that is outside the scope of 230.

[mirimir]

This has been my understanding for many years. And so it's ironic how increased moderation to increase marketability brings immunity of service providers under 230 into question.

It's almost like 230 has been a bait and switch ploy. With 230, mega social media corps developed, with nothing else to protect them against liability.

But instead, we could have had decentralized systems that made liability impossible. And if stuff like this goes forward, maybe we can. Someday, anyway.

[dragonwriter]

> This has been my understanding for many years. And so it's ironic how increased moderation to increase marketability brings immunity of service providers under 230 into question.

Especially since allowing providers to increase (automated and best-effort, but not comprehensive human) moderation withour incurring general liability for content was the explicit and overt justification for Section 230, because the kind of complete human editorial control expected of publishers in traditional media was viewed as preventing scalable systems on the internet (where traditional media had other scaling limits that make content liability far from the limiting factor.)

> But instead, we could have had decentralized systems that made liability impossible.

Decentralized systems probably wouldn't have made liability impossible, just ineffective at it's objectives, because no matter how many operators were ruined by liability. the content would still thrive and you'd never hit enough operators.

[mirimir]

Yes, decentralized wouldn't be enough. I'd want anonymously decentralized.

[schoen]

Do you think many users will be willing to adjust to the complexity and unpredictability of that environment compared to what they experience now? Just because it's mostly way better in terms of issues of freedom and power?

[mirimir]

Sure, I think. You just default client apps to all mainstream filters. So new users see nothing alarming. At least, until they start tweaking.

[schoen]

I said recently in another thread about this (https://news.ycombinator.com/item?id=22352166) that

> I don't tend to think tinkering with intermediaries' incentives about content is the thing that will get us there, because there are so many other practical advantages that people have perceived in the more centralized services.

Do you think otherwise about that?

[mirimir]

I do understand why loss of §230 protection is dangerous, both directly and as precedent. Threatened intermediaries would just do whatever needed to avoid liability. And incentives for alternatives would likely remain inadequate.

Also, I'm not advocating any "tinkering". I'm just not optimistic about relying long term on legal protection. Even for the US, it's been fragile, and not anywhere near effective enough.

I was just being wistful, speculating that, without §230 protection, we'd have ended up with a system that didn't need such protections. With intermediaries that were totally isolated from content, with absolute anonymity for users. That is what I was imagining in the mid 90s.

I can imagine more or less centralized services that could handle content which was otherwise end-to-end encrypted and anonymized. And could make money doing it, as VPN providers get paid by Orchid users via an Etherium-based currency.

Users would have the tools to filter what they see. But nothing could be removed, and nobody could be prosecuted (or persecuted).

How we might get there from current social media, I have no clue.

[schoen]

Thanks for those clarifications. That makes sense to me.