Corynne McSherry

It's not the only reason, but it helped.

But I want to answer Jad's question, because I think what you're asking is for an example of why I'm worried about how the moderation happens.

Yeah, OK.

So the way that it works now and the way that it's likely to continue to work is that the social media companies employ a combination of humans and mostly algorithms to try to figure out what's bad speech and what's good speech.

And they mess it up.

So they'll end up taking down this statement, all white people are racist, as an example of hate speech, but they won't take down

If you might show the previous one, this from a congressman who said, not a single radicalized Islamic suspect should be granted any measure of quarter, et cetera, et cetera.

Nasty stuff, right?

They can't tell the difference.

And that's what happens.

And there's a hat tip to ProPublica.

I hope you guys are all ProPublica supporters and fans because they're great.

They did a detailed study to look at Facebook's policies.

And they found out that, among other things, they're training their moderators to, in some instances, protect white men over black children.

That's where we are right now.

That's what we want to endorse.

That's what we want to encourage.

I don't think so.

What we have, where we are right now, is thousands of accounts are being suspended every day.

Let's just say a relatively small percentage of those are for perfectly lawful speech.