Sacks commented on the question raised again (because governments are trying to charge platform owners again, citing terrorism and other crimes ‘hosted’ on their platform, in an attempt most understand to be a simple grab for control). Chamath, who doesn't have a journalist background, raised the algos-are-like-publishers argument. And the 230 law. Algos are better every day at doing one-to-one tailoring of information, reacting to how you interact with the platform's content.
Algos, Sacks said, simply show you more of what you interact with, assuming that if you interact with something, and it takes effort to interact, you want to see more of that. A couple years ago it was found that people were complaining about being shown outrage-content, but when they looked into it they found that the users had been sharing that content with friends in their outrage at it. The algo just showed them more of it. There's no way for the algo (or anything, maybe) to know when you're interacting with content but you don't want to see more of it. ‘We are currently not smart enough to realize that,’ tweeted Musk a couple days ago.
An editor has an editorial viewpoint, message, motivation.
‘Consumers want to have stuff that incites emotion. There’s a reason horror movies do well, as well as romantic comedies and action films,' said Friedburg. The more likely it's emotive, the more likely you are to want something like that again, he said. What would happen if algo's starting showing things that were less inciting? Would the platform just lose to competitors?
Algo's also function with momentum, decay. If you start watching 15 minutes of something, it'll show you lots of that for a while.
"Founder Mode," DOJ alleges Russian podcast op, Kamala flips proposals, Tech loses Section 230? (youtube.com)