The following podcast contains advertising to access an ad-free version of the LawFair

podcast become a material supporter of LawFair at patreon.com slash LawFair.

That's patreon.com slash LawFair.

Also, check out LawFair's other podcast offerings, rational security, chatter, LawFair

no bull and the aftermath.

We know what the primary tools in the toolbox are, but we don't know how relatively effective

they are and what the trade-offs are.

And in the absence of some ability to evaluate that, it's kind of like taking shots in the

dark and you can see it with the responses the platforms are making to various instances

of misinformation and disinformation.

The part of the reason why these responses feel so scatter shot, some posts get a contextual

warning, some get a little more, sometimes the post is removed, sometimes the account is

removed, sometimes it's a temporary removal is because we don't really have a principled

basis for making good judgments about which intervention is the right fit for which kind

of problematic content.

I'm Alan Rosenstein and this is the LawFair podcast August 10, 2021.

The spread of misinformation is one of the biggest challenges facing social media platforms.

A standard approach is to label suspicious posts or links so as to warn users that what

they're quote unquote engaging with is not reputable.

But warnings, despite their wide use, have been proven to be particularly successful.

So what's the social media platform to do?

Two Princeton University computer scientists, Ben Kaiser, a PhD student and Professor Jonathan

...