State sponsored radicalism

From The Economist:

Gonzalez v Google was brought by relatives of a woman killed by Islamic State in a terrorist attack in Paris in 2015. They say that YouTube, Google’s video platform, shares blame for her death as its algorithm fed radicalising clips to potential IS recruits. Google contends that without the protection of Section 230 sites will lose the capacity to help users find “needles in humanity’s largest haystack”.

From another perspective, Google is arguing that its business model has no hope of being profitable without the government shielding Google, and other tech companies, from the negative impact of their products. Other industries don’t get this sort of protection.

Perhaps this means that the entire concept of YouTube shouldn’t exist, as there’s no way to allow people to watch and upload massive amounts of video for free without having to resort to prodding viewers to ever more extreme forms of “engagement”.

It’s interesting that podcasts as a still largely decentralized thing are thriving. If you want to upload a podcast, you have to pay for the hosting. There isn’t a need for dodgy engagement schemes as the platforms aren’t making more money by endless engagement. Yes, individual podcasts can make more money by being dramatic, but it doesn’t look like this leads to the sorts of extremism as when platforms incite divisive content.

That said, most podcasts are rubbish. But that’s fine. Beauty is in the eye of the beholder.

What is good about them, is that podcasts show that decentralization may just be the key to content moderation. Without the huge financial incentives for engagement by platforms, it’s a lot harder to stir the pot.