Hacker News Reader

Comment by charlesism | Hacker News Reader

There's plenty of positives to Reddit. The trouble is that the prevailing philosophy behind Reddit (and FB, Youtube, etc) doesn't accomplish what everyone thought it would. We have been operating under the assumption that "filter bubbles are bad" and that private businesses should uphold "free speech" the way the judicial system does in the public realm.

In practice, it looks as though online conversations aren't constructive unless the participants share a certain amount of common ground (eg: "first principles" like whether the strong should dominate the weak, or whether all "MSM" are part of a conspiracy and untrustworthy). I think in the future social media startups will have a "point of view" and not try to cater to everyone.

View on HN

>We have been operating under the assumption that "filter bubbles are bad"

The entire premise of Reddit is that filter bubbles are the solution. You get to create bubbles for your own views called 'subreddits' where you can moderate out any wrong-think until the remaining subscribers reinforce the correct-think with their upvote/downvote mechanisms.

A system that operated under the assumption that "filter bubbles are bad" would not allow its members to upvote/downvote each others' content so simply.

I worded that poorly. I mean filter bubbles are the right idea, but don't go far enough. A site should should hold specific enough values that its users can debate each other in good faith, and so that admins can stamp out (what they consider to be) toxic behavior without shocking half their users.

>specific enough values that its users can debate each other in good faith

you agree with me? great. wonderful debate thanks guys

Just because two people are, say physicists or atheists or believers in liberal democracy, doesn't mean they agree about everything (or even much at all).

The past few years make me think that without sufficient common ground, rational debate is impossible:

•Alice: "Rosemary's Baby is the best film ever." •Bob: "Devil woman! Satan is speaking through you. But you can't fool me!"

•Alice: "My father just died. Everyone, give your dad a hug. You'll miss him when he's gone." •Bob: "Cry me a river, loser! Cow farts. Cow farts. Hitler was right!"

•Alice: "I didn't vote for Trump, but I am glad he's talking to North Korea" •Bob: "Of course you are, Russian bot! I hope your handler is happy"

•Alice: "I used to be a cop, and this video is way over the line." •Bob: "Antifa scum! How much does Soros pay you to write this BLM propaganda?"

•Alice: "Just a reminder that you can get a free vaccination today, if you live in Charleston." •Bob: "Great, free autism everybody! Bet you'll trot out your MSM 'facts' now. Science is the new religion."

The web is so full it's bursting with "debates" like these. They are not worth having. They do not change anyone's mind. They cause people to dig in. They just further divide society.

In all of those cases, Bob is being a dick. A rule to not be a dick is pretty straightforward and can be enforced by moderators.

Moderators don't scale very well though. Those filter bubbles are really just letting the users do their own moderation. This reduces the scope of the site moderators job down to enforcing the rules.

If you try to do everything through the global site moderators you immediately get pulled into arguments like this: User 1: You need to ban User 2 because he is a racist. User 2: I only said that User 1's views are not supported by evidence. User 1: My views are well supported by my life experiences, and there is no good evidence because the government outlawed funding for research into them.

It's a no win situation, and neither side is going to change their mind and there's no clear winner. They're going to bicker and raise your noise floor constantly though. It's much better for site health if they can be isolated from one another, and better for your moderator's health if they do it automatically by themselves.

This is the messy messy world of real life.

Maybe we could help determine opinion from referencial material so that we can speed up verbal communication rather than trying to evolve it.

YouTube and Wikipedia working together are good strides. LinkedIn offers suggestions on how to respond to people, then the new Gmail came out with that same feature, I hope Airbnb does too.

Maybe, when using curse words/slang AI could suggest something more intelligible?

If we connected our brains to an emotional evaluator, we could include all the necessary emoji automatically, so we don't accidentally forget that winky face when point out an irony or satire (satire; often mistaken for sarcasm due to different emotional attachments).

We need a way to determine if the debate responses are rational (not currently possible on free text without super non-existent AI or human moderators).

Trying to find common ground just eliminates useful debates because your solution ends up being "put the people who agree with each other together". Also, people having common ground doesn't mean they will behave rationally when they run into differences.

I would rather debate socialism with someone rational who believes in it rather than a strawman portrayed by someone else in the capitalist bucket I get thrown into.

Notably, by this standard Reddit isn't a site. Each subreddit, or cluster of subs with common moderators, is functionally a distinct site.

Obviously this isn't absolute. The site has a common structure for debate, the default subreddits have common traits (and common troll/bad-faith factions), and the media perception is of a unified site.

But the functioning of /changemyview, /askhistorians, and other highly structured subs is vastly different from "the reddit experience", and lots of other subs with less structure still have mod-enforced debate boundaries. As an example - the sub for "late capitalism" memes bans defenses of capitalism or debates on the merits of socialism to create shared values for discussion, while linking to /debatecommunism to provide space for a different debate.

Honestly, this is my biggest frustration with talk about reddit as "an alt-right site" or predictions that it will die out from ideological conflict. It's not Facebook, where things pass arbitrary distances along a graph. It's not Tumblr, where the basic communication structure is "here, talk to people who hate you". It's a siloed system where a huge fraction of users never experience any of what's popularly described as 'Reddit'.

>As an example - the sub for "late capitalism" memes bans defenses of capitalism or debates on the merits of socialism

Tangential: I honestly don't understand why someone would willfully participate in such obvious propaganda (even if you 100% believed eliminating capitalism is the solution). It's willfully burying your head in the sand to pretend that your ideology that would replace the modern western world isn't open to criticism.

The best argument I can give about "open to criticism" is that it's meant as space to vent. These same people might be over in /debatecommunism, but they want a place to laugh about what they see as obvious flaws and evils without an endless string of debates.

That's an impulse I can understand, but the exposure-to-propaganda part I can't get my head around as easily. Even political cartoons/memes that align with my positions mostly just feel slimy these days - like they're not about humor or solidarity, but pushing people to take the angriest, most simplistic takes possible on an issue. Most of /latestagecapitalism or /libertarianmeme or anything else is completely misleading no matter what politics you subscribe to.

Reply on news.ycombinator.com