I've kind of wondered if flattening the world isn't the root problem with social media. There are very few people I encounter IRL that I disagree too much with. We do disagree, but everyone kind of understands each other to some degree. In contrast, my values and the values of someone in Germany (as an example), might be less aligned. We frame so much of this as either echo chambers (we work in the set of people who agree) or ML targeted rage bait. Maybe the best algorithm is Californians mostly engaging with Californians with a spectrum of viewpoints that naturally exists from the urban rural divide.
Internet in the 90s taught me that I have more in common with like-minded people around the world than with the people I meet in daily life. But then ads became the primary business model for online communities, and the need to drive engagement turned most of them toxic.
And then I moved to one of the least affordable areas in the US, which is also a college town and a tourist destination and has huge issues with homelessness and property crime. Here you don't have to use social media anymore. If you want conflict, you can just talk to your neighbors about local issues. Your financial interests and ideas about the future of the community are guaranteed to be in conflict with many of them.
It's often not somebody from Germany, but a troll farm of humans or LLMs controlled by an intelligence agency with the explicit goal of making you miserable and disenchanted. This has been ongoing since at least 2020 across all social media platforms.
The platform itself does a wonderful job of promoting outrageous/divisive content since this yields much more "engagement" (which is how they make money) than feel-good, innocent content.
This disgusting business model itself is the problem, not some malicious foreign actors (though they no doubt can take advantage of the free exposure the platform will give them if their content "engages" enough people).
I went to school with some Russians so I was often able to pick out troll profiles based on slight grammar mistakes and word choice. I imagine LLMs are going to make this harder.
I'm a Russian and sometimes I think they did evolve to levels that are harder to detect now. Or I am just more paranoid from seeing the trolls and over the place and knowing about LLMs. Or there are more people who got converted by those trolls into zombies. Or all 4.
What is the difference between that and masses of years using downvotes or reporting to make everyone they disagree with feel miserable and disenchanted? I feel like the notion of a troll farm is exaggerated while the influence of regular users and moderators in manipulating others is ignored.
I don't think so. Have a read of these articles and get back to me.
Beyond examples, it would be highly surprising if they weren't doing it. I think the burden of proof is on the person saying that they wouldn't do the thing that any sensible realist actor would do to a rival. That would be a level of bizarre incompetence that is too much to fathom.
I looked at half of them and found nothing about how much stuff people see on social media is this. One apparently had its content seen by 140 million Americans. So what does that mean from the point of view of an individual? Is it most of what they saw? 10%? 1%? 0.0? I have no idea and neither do you.
It's a popular misinformation technique of news companies to do this - planting a false idea in the heads of their enthusiastic audience using truthful information presented in a blustery way that leads people to believe something it doesn't actually say.
So you agree it's not a "conspiracy theory", then? And your response to troll farms serving impressions to half the country is to argue that I don't know exactly what percentage of social media content is generated by troll farms? You are being obstinate.
"serving impressions to half the country" you're still confused about the meaning of the information. Average people probably see hundreds or thousands of social media posts in a month so if only one of them is from the troll farm, that's essentially none.
You said it's often a troll farm or LLMs controlled by an intelligence agency. I don't see any evidence for that. Unless you take the pedantic meaning of "often" to be "not never but possibly hardly ever".
I'm not denying that it happens, but that it's as significant as you presented it to be.
The "problem" (really, what makes it great) with social media is that it isn't real. You can adopt a different "persona" and explore ideas that are incompatible with the person you actually are, making it a fantastic way to learn. The, uh, "no so with it" crowd sometimes becomes confused by what it is, but such is life. Nothing is perfect.
The real experience is also there to explore, and is worth exploring, but there is also value in trying something else once in a while, and that is what social media offers. After all, if it were just a digital duplication of outside, why wouldn't you just go outside?
