[KaiserPro]

> You can be protected by safe harbour provisions, or you can editorialise your content.

Aha, now this is an interesting distinction. I'm not an expert in this, as you might imagine, but what counts as editorialising?

To my naive eyes, having an algorithm that re-arranges posts, or injects new subjects seems like editorialising to me.

[traverseda]

I'm also not a lawyer, I was making that as a more vague moral distinction on the topic of free speech and accountability.

For practical reasons I think those algorithms are absolutely necessary. We need spam filters. A good line to draw would be "bring your own algorithm". A technical challenge to be sure, bit breaking up social media backend providers and content filtering seems like one of the only safe ways to allow these massive platforms to exist.

The algorithm can be just "Dan filters out spam".

[mystraline]

Even spam filters are problematic.

At first, its just unsolicited commercial crap.

Then its non-corporate allowed unsolicited commercial crap.

Then its 'hide commercial crap in posts to deceive'.

Then its 'fuck over screen readers by aligning everything weird like FB to prevent finding commercial crap'

Then its "hey we can add these other non-spam categories (like Palestine) to silence them".

[thaumasiotes]

> Aha, now this is an interesting distinction.

It's nothing new; the entire point of §230 is to provide protection to platforms that editorialize their content. Without editorializing, you have immunity anyway.