3 ms·
All this talk of changing S.230 and yet nobody has legally tested what the current situation is with regards to algorithmic editorialization. If Bob posts somet
by mindslight 5y ago
All this talk of changing S.230 and yet nobody has legally tested what the current situation is with regards to algorithmic editorialization. If Bob posts something that mechanically goes to all his friends, then S.230 straightforwardly applies. But if Bob posts something that is merely feedstock for Faceboot (etc) to possibly decide to pass on to some of his friends, then it seems that the middleman has a large part in the speech. You can't join in with someone singing a song, and then disclaim responsibility for its contents.
If Mary's friends all post a bunch of Covid denialism, and Mary gets sick and dies, it's the same situation as if they had called her up and spoke to her directly. If a small portion of Mary's friends post Covid denialism, and then Faceboot determines that Mary is susceptible to the narratives of Covid denialism and will "engage" more with their site if they repeat these messages to her, then Faceboot is directly responsible when Mary gets sick and dies.
I think that is close to the limits of what is possible to do with respect to the first amendment, and possibly close to the current legal situation depending on how a lawsuit with such theories fared.
The only plausible approach I've seen to breaking up the power of Big Tech is through a US GDPR and mandating open interoperability APIs. The data protection is so that individuals can choose not to be subjects of these companies, and the open APIs are so that companies have to compete for business rather than coasting on Metcalfe's law. Every other proposal I've seen is basically a way for the government to control Big Tech's power for its own ends, rather than dispersing it so that such concentrated power doesn't exist in the first place.