Comment by tensor
5 months ago
There is a difference between free flow of information and propaganda. Much like how monopolies can destroy free markets, unchecked propaganda can bury information by swamping it with a data monoculture.
I think you could make a reasonable argument that the algorithms that distort social media feeds actually impede the free flow of information.
> Much like how monopolies can destroy free markets, unchecked propaganda can bury information by swamping it with a data monoculture.
The fundamental problem here is exactly that.
We could have social media that no central entity controls, i.e. it works like the web and RSS instead of like Facebook. There are a billion feeds, every single account is a feed, but you subscribe to thousands of them at most. And then, most importantly, those feeds you subscribe to get sorted on the client.
Which means there are no ads, because nobody really wants ads, and so their user agent doesn't show them any. And that's the source of the existing incentive for the monopolist in control of the feed to fill it with rage bait, which means that goes away.
The cost is that you either need a P2P system that actually works or people who want to post a normal amount of stuff to social media need to pay $5 for hosting (compare this to what people currently pay for phone service). But maybe that's worth it.
>We could have social media that no central entity controls, i.e. it works like the web and RSS instead of like Facebook. There are a billion feeds, every single account is a feed, but you subscribe to thousands of them at most. And then, most importantly, those feeds you subscribe to get sorted on the client.
The Fediverse[1] with ActivityPub[0]?
[0] https://activitypub.rocks/
[1] https://fediverse.party/
Something along those lines, but you need it to be architectured in such a way that no organization can capture the network effect in order to set up a choke point. You need all moderation to be applied on the client, or you'll have large servers doing things like banning everyone from new/small independent servers by default so that people have to sign up with them instead. The protocol needs to make that impossible or the long-term consequences are predictable.
3 replies →
There is no generally accepted definition of propaganda. One person's propaganda is another person's accurate information. I don't trust politicians or social media employees to make that distinction.
There is definitely videos that are propaganda.
Like those low quality AI video about Trump or Biden, saying things that didn't happened. Anyone with critical thinking knows that those are either propaganda or engagement farming
Or they're just humorous videos meant to entertain and not be taken seriously. Or they are meant to poke fun of the politician, e.g. clearly politically motivated speech, literally propaganda, but aren't meant to be taken as authentic recordings and deception isn't the intent.
Sometimes it's clearly one and not the other, but it isn't always clear.
2 replies →
What you think is propaganda is irrelevant. When you let people unnaturally amplify information by paying to have it forced into someone’s feed that is distorting the free flow of information.
Employees choose what you see every day you use most social media.
Congrats! You are 99% of the way to understanding it. Now you just have to realize that "whoever is in charge" might or might not have your best interests at heart, government or private.
Anyone who has the power to deny you information absolutely has more power than those who can swamp out good information with bad. It's a subtle difference yes, but it's real.
5 replies →
OK, but that's an argument against advertising, and maybe against dishonest manipulation of ranking systems.
It's not an argument for banning doctors from YouTube for having the wrong opinions on public health policy.
> distorting the free flow of information
There is no free flow of information. Never was. YouTube and FB and Google saying "oh it's the algorithm" is complete BS. It always manipulated, boosting whoever they feel fit.
And propaganda by definition isn’t false information. Propaganda can be factual as well.
So many people have just given up on the very idea of coherent reality? Of correspondence? Of grounding?
Why? No one actually lives like that when you watch their behavior in the real world.
It's not even post modernism, it's straight up nihilism masquerading as whatever is trendy to say online.
These people accuse every one of bias while ignoring that there position comes from a place of such extreme biased it irrationally, presuppositionaly rejects the possibility of true facts in their chosen, arbitrary cut outs. It's special pleading as a lifestyle.
It's very easy to observe, model, simulate, any node based computer networks that allow for coherent and well formed data with high correspondence, and very easy to see networks destroyed by noise and data drift.
We have this empirically observed in real networks, it's pragmatic and why the internet and other complex systems run. People rely on real network systems and the observed facts of how they succeed or fail then try to undercut those hard won truths from a place of utter ignorance. While relying on them! It's absurd ideological parasitism, they deny the value of the things the demonstrably value just by posting! Just the silliest form of performative contradiction.
I don't get it. Fact are facts. A thing can be objectively true in what for us is a linear global frame. The log is the log.
Wikipedia and federated text content should never be banned, logs and timelines, data etc... but memes and other primarily emotive media is case by case, I don't see their value. I don't see the value in allowing people to present unprovable or demonstrably false data using a dogmatically, confidentally true narrative.
I mean present whatever you want but mark it as interpretation or low confidence interval vs multiple verified sources with a paper trail.
Data quality, grounding and correspondence can be measured. It takes time though for validation to occur, it's far easier to ignore those traits and just generate infinite untruth and ungrounded data.
Why do people prop up infinite noise generation as if it was a virtue? As if noise and signal epistemically can't be distinguished ever? I always see these arguments online by people who don't live that way at all in any pragmatic sense. Whether it's flat earthers or any other group who rejects the possibility of grounded facts.
Interpretation is different, but so is the intentional destruction of a shared meaning space by turning every little word into a shibboleth.
People are intentionally destroying the ability to even negotiate connections to establish communication channels.
Infinite noise leads to runaway network failure and in human systems the inevitably of violence. I for one don't like to see people die because the system has destroyed message passing via attentional ddos.
Fortunately your biased opinion about what information has value is utterly worthless and will have zero impact on public policy. Idealized mathematical models of computer networks have no relevance to politics or freedom of expression in the real world.
There isn’t. Yet, everybody knows what I mean under “propaganda against immigration” (just somebody would discredit it, somebody would fight for it), and nobody claims that the Hungarian government’s “information campaign” about migrants is not fascist propaganda (except the government, obviously, but not even their followers deny it). So, yes, the edges are blurred, yet we can clearly identify some propaganda.
Also accurate information (like here is 10 videos about black killing whites) with distorted statistics (there is twice as much white on black murder) is still propaganda. But these are difficult to identify, since they clearly affect almost the whole population. Not many people even tried to fight against it. Especially because the propaganda’s message is created by you. // The example is fiction - but the direction exists, just look on Kirk’s twitter for example -, I have no idea about the exact numbers off the top of my head
Propaganda wouldn't be such a problem if content wasn't dictated by a handful of corporations, and us people weren't so unbelievably gullible.
indeed, didn't YT ban a bunch of RT employees for undisclosed ties? I bet those will be coming back.
Oh, but can you make an argument that the government, pressuring megacorporations with information monopolies to ban things they deem misinformation, is a good thing and makes things better?
Because that's the argument you need to be making here.
You don't even need to make the argument. Go copy paste some top HN comments on this issue from around the time the actions we're discussing youtube reversing happened.
I think those arguments sound especially bad today, actually. They got the suppression they wanted, but it did not give the outcome they wanted.
Not really. You can argue that the government should have the right to request content moderation from private platforms and that private platforms should have the right to decline those requests. There are countless good reasons for both sides of that.
In fact, this is the reality we have always had, even under Biden. This stuff went to court. They found no evidence of threats against the platforms, the platforms didn't claim they were threatened, and no platform said anything other than they maintained independent discretion for their decisions. Even Twitter's lawyers testified under oath that the government never coerced action from them.
Even in the actual letter from YouTube, they affirm again that they made their decisions independently: "While the Company continued to develop and enforce its policies independently, Biden Administration officials continued to press the company to remove non-violative user-generated content."
So where does "to press" land on the spectrum between requesting action and coercion? Well, one key variable would be the presence of some type of threat. Not a single platform has argued they were threatened either implicitly or explicitly. Courts haven't found evidence of threats. Many requests were declined and none produced any sort of retaliation.
Here's a threat the government might use to coerce a platform's behavior: a constant stream of subpoenas! Well, wouldn't you know it, that's exactly what produced the memo FTA.[1]
Why hasn't Jim Jordan just released the evidence of Google being coerced into these decisions? He has dozens if not hundreds of hours of filmed testimony from decision-makers at these companies he refuses to release. Presumably because, like in every other case that has actually gone to court, the evidence doesn't exist!
[1] https://www.politico.com/live-updates/2025/03/06/congress/ji...
The key problem with the government "requesting" a company do something is that the government has nigh infinite unrelated decisions that can be used to apply pressure to that company.
It's unreasonable to expect some portion of the executive branch to reliably act counter to the President's stated goals, even if they would otherwise have.
And that opportunity for perversion of good governance (read: making decisions objectively) is exactly why the government shouldn't request companies censor or speak in certain ways, ever.
If there are extenuating circumstances (e.g. a public health crisis), then there need to be EXTREMELY high firewalls built between the part of the government "requesting" and everyone else (and the President should stay out of it).
3 replies →