I wonder if social media had the potential to be a force of good or if it was doomed from the start. I'd like to run an AB experiment where social media companies didn't optimize for engagement. And maybe an AB experiment where they tried to optimize for healthy usage, even if it harmed engagement.
This is just a theory but my fear is that social media that's optimized for "healthy usage" probably looked more like the forum and blog culture that social media killed.
Optimizing for engagement means you basically have a genetic algorithm on your hands for surfacing the content with the most "viral-potential." Eventually that stuff eats the healthy parts of the internet because people inevitably talk about the viral stuff that's happening, which means your healthy-use forum is nonetheless revolving around the conversation in the viral centers.
From there it's a matter of time before people start going directly to the viral source to keep up with the context and conversation. And once they're there, because it optimizes for engagement, it crowds out their use of everything else.
So there's a natural selective pressure here. Optimizing for engagement/addiction gets you a network effect that leads to overshadowing any other type of socializing. Unless there's some mechanism to actively select against virality and engagement they will naturally rise to the top even independent of ad-impression incentives.