You know what, though? I kind of think you don't even
need bots for the "majority opinion" to get widely skewed.
For example, you can find an absolute ballistic lunatic take on Twitter that has 10,000 retweets basically trivially.
Of course, some of that is bots. Maybe someone paid for engagements. It's easy to do so, so it's naturally what a lot of people assume.
Even if it wasn't mostly bots... 10,000 people? It feels like a lot, enough to make us feel like that opinion is at least somewhat validated socially. But in reality, it's just a number. Would we even care who agreed if we knew what types of people they were?
But even if we did still care, all this really tells us is that 10,000 people hit a button. Why? A lot of big creators could say virtually anything and immediately get social validation in return, I strongly doubt that the majority of people interacting really have an actual strong conviction about whatever position is being expressed.
But even if they all really did hold a strong conviction... how many people is 10,000 really? Twitter has several hundred million users. Around 100,000,000 active U.S. users each month. It is still kind of impressive in a sense that 10,000 people saw a post and decided to click a button I guess, but I think our intuition is really broken when it comes to large enough numbers of people; it does not suggest that most people, or even a large minority of people, actually agree with it.
But what about the posts we don't see? The average Twitter user has 770 followers, but that number is highly skewed by very large accounts... One metric I found online, though I couldn't find a good source, suggests that 0.06% of users have over 1,000 followers. There are accounts with millions of followers. Needless to say, the posts that find the most visibility on Twitter are largely curated by a very elite group of users... naturally, the majority of "popular" posts you're going to see are not just random posts that happened to catch on. They're mostly posts that a large account boosted! Naturally this is going to cause all sorts of problems.
You could go on and on and on. The social media Internet is basically a giant false plurality machine. There doesn't need to be bots. There doesn't need to be troll farms. There doesn't need to even be bad actors, malice, disinformation campaigns.
And it's silly, because our brains are obsessed with what "most people" think, and I think there are some rational, logical reasons to care about this, but I don't believe in earnest that this is mostly coming from a rational place. People desperately want to feel belonging, and to feel validated in their opinions. How many people here on Hacker News reply then come back a few minutes later to check and see if the post got upvoted or downvoted? Who doesn't have at least a little bit of a "dopamine rush" when they have a post that gets hugely upvoted on Reddit or Hacker News, or reposted a lot on Twitter/Mastodon/etc.?
Yet, even though we know about this, we don't then earnestly take this into account when we see "popular" opinions. You ever notice how obsessed the social media Internet has become with condemning social taboos? Seems pretty straight-forward to me, it's an easy way to get that "people agree with me" dopamine hit. I'm not even saying that people intentionally do this, either. People unknowingly shape their behavior around what they think will get them that positive attention. I sincerely doubt I am immune from it, I'm sure all of my Hacker News posts wreak of HN-specific self-censorship and intricate codeswitching even though I really try not to do that sort of thing. Hell, scrutinize this paragraph: "Oh, 'social taboos'? Afraid to just say the one we're all thinking of?"
Social media isn't even alone, false pluralities absolutely spread in the real world, too, I just think that social media is really good at doing it faster and more intense than ever before.
Honestly, I just think we care too much about what "everyone" thinks. Imagine the laws if they were based on what the majority of people think they should be. Imagine Wikipedia if the rule for resolving conflicts was a poll of what people think should go in the articles. Even if we really could know what the majority of people truly think, a lot of the time, I'm not sure we should do anything with that information. Do most people really think hard enough about problems to really have an informed opinion on most issues?
I think, instead of focusing on what most people think, we should just simply always seek to do the right thing, and seek to be as "correct" as possible in a world without absolute objective truth. I'm not a politics person, but I think this sort of thing is exactly what abstractions like having representatives is good for... it's kind of unfortunate that in the real world, these systems can wind up being corrupted to the point of being a net negative.
And I think that means we should just do our best to ditch websites like Twitter that are basically as unproductive as possible here. Maybe some Twitter alternatives can do a little better here, particularly ones that allow for smaller groups and communities to have their own spaces, but really I just think the entire "model" is not very good. For all of the faults of structured discussion forums like Hacker News, unstructured social media just doesn't seem to work very well for much other than soaking up attention and moving tons of advertising dollars.