2w ago

Europe's race to rearm is pointless if its adversaries are waging war online

www.theguardian.com

Europe's race to rearm is pointless if its adversaries are waging war online | Johnny Ryan

This opinion piece argues that Europe should "shut down recommender algorithms" of the big US social media platforms - Facebook, X, Instagram, etc. - because the author believes that these algorithms are undermining European democracy.

The most obvious example of such an algorithm is on X, where Musk can manipulate the algorithm to boost European far-right parties, like AfD. But the author argues that other social media CEOs, like Zuck, are beholden to Trump's anti-liberal agenda - for example, Trump "openly threatened to throw Mark Zuckerberg in jail for the rest of his life". Therefore: "It is reasonable to assume that tech oligarchs will do what [Trump] tells them", which may include the Trump administration pressuring US social media companies to recommend more right-wing content.

So the author says: "The EU must immediately switch off the tech companies’ algorithms on its soil, at least until they are proven safe for democracy". Do you agree with that?

29 comments

They’re right. Algorithms are arguably responsible for a lot of the political and social issues that we’re experiencing right now. They’re literally designed to divide us.
What I mean by that, is that these algorithms were built to maximise engagement, so they would promote content based on how engaged (how many ‘clicks’) users would get, but without any other context about the content (eg. Factuality, impartiality etc). It turns out that content that triggers intense emotional responses are the most engaging, so these algorithms would promote emotionally engaging content, and nothing gets people more engaged than outrage and anger. So more and more algorithms would promote content that made people angry and outraged, and when companies and political groups realised this, they latched on to their new found ability to manufacture outrage for their own benefit.
That’s where we are now. Algorithmically driven social media is essentially a tool for collecting personal data, and dividing its users for political gain.
We spend so much time worrying about the impact of AI, that we don’t realise the impact of these things that are already causing massive amounts of damage.
- It's important to note that this isn't a hunch. This has been studied and proven. We know this is happening and nobody is doing anything on a political level to stop it.
  
  Thanks for this.
- Agreed. Algorithms, bots and fake news work together to create and perpetuate harmful narratives.
  I think the DMA has the right idea and it should be extended. Any platform with a large user base should be heavily regulated and controlled.
  We can still have mostly unregulated and uncensored free speech on small platforms, since those aren't harmful for our democracy.

The EU should be promoting the shit out of opensource platforms and dumping millions into them both for content and development.
Subsidise media made by companies wholely owned by Europeans and add a clause forcing them to distribute on European platforms that host in the EU and are made available all across the EU without geoblocking. Massively promote those EU platforms on foreign social media (which is pretty much all we have at the moment) e.g pay for ads on youtube, insta, tiktok, and so on to "watch this on peertube" or whatever else EUropean we have.
Also, what a shit headline. It's not pointless, it's one of the many things Europe has to do. Stop expecting perfection. Doing something is better than nothing, but you have to start somewhere instead of planning for perfection and never starting.
- The Fediverse isn’t immune to propaganda and bots. It might even be an easier as in cheaper target than Twitter or TikTok.
  Creating a bunch of superficially legitimate instances and using them to upvote stuff you like is straightforward.
  
  I think it'll be more difficult to fake and account of the European commission for example as the domain cannot be captured. An official instance or person will have @officialdomain.eu or something. And mastodon doesn't have downvotes.
  As for bot instances, those could be blocked by other instances. It wouldn't surprise me if government instances implemented a blocklist or a functionality to only allow communication or interactions with approved instances. There is a lot that governments will be able to do, but with Xitter, Facebook, and so on, they first have to debate thing, tell the site to do something, go through the appeals process, and so on. Owning the instances would allow them to take action much more quickly.

I've been following the German election campaign very closely and the only time social media was even mentioned was when a young lady in one of the town hall formats asked if there were any plans to deal with the danger of social media in the context of impact on young people's development. The question was put to Alice Weidel. Her answer was basically “No idea. I haven't dealt with it yet”.
That was it. Otherwise, the topic wasn't even mentioned in any discussion round.
But every single discussion round opened with “the topic that every German is talking about”: migration migration migration migration. Wonder why everybody is talking about that?
I would like to emphasize this sentence from the article because it is simply true:
What we do in these next few months will determine whether Europe's liberal democracy survives or is lost for ever.
Something has to happen NOW, but nobody is talking about it. It's really exasperating.
- The question was put to Alice Weidel. Her answer was basically “No idea. I haven’t dealt with it yet”.
  Classic fascist answer: The AfD heavily used the absence of other parties on particularly TikTok to propagandise, they would structure their parliament speeches just for soundbites on TikTok. Once spotted Die Linke started to strategically engage them also there, the other parties still don't have coherent strategies. Individual politicians are doing things, but no proper campaign.
  Which kinda isn't a new phenomenon e.g. the SPD had tons of newspapers back in the days and pretty much sold all of them. Some kept their edge (e.g. the Hamburger MoPo, not to be confused with the Berlin Morgenpost), but most washed out to standard liberal press gray. CDU, FDP? Exists by grace of the Springer press, they're on the short end of the lever. Greens, well, the taz exists but who reads the taz. The MoPo works because despite being far from gutter press, it's still a tabloid: You gotta have those sportsball results. A paper to read in the metro, talk about in a bar and leave at a construction site, not a coffee house.

Social media algorithms are bad, but quick reminder that Hitler and Mussolini didn't need algorithms to overthrow democracy. The root problem here is either capitalism or neoliberalism depending on how radical you wanna get, and at the very least the latter has to go yesterday.
- The Nazis owned and operated their own newspapers, nearly from the start, to get their propaganda out.
  
  They took over existing publications as well. Still, I'm all for reigning in the social media networks, because they're basically editorialising through their moderation (like that time Musk decided to boost his own posts so much he ended up on everyone's 'for you' page).
  If your moderation is basically taking on the role of an editor, you should be held responsible for any libellous content/ hate speech you boost in this fashion.
- The old time fascists used the new media of their day to the utmost. Their propaganda was so effective, it still echoes today. Fascism itself is focused on aesthetics, appearance, and performance. All of these are forms of propaganda.
  
  could you elaborate about facism being focused on aesthetics, appearance, and performance?

- it should be like the fediverse!

I absolutely agree with that. The algorithms are one of the significant promoters of destabilization, propaganda, disorientation, and manipulation, by platforms and by other parties making use of them.
It is essential for the EU to regulate them. Through transparency and requirements, and adequate consequences.
It was one of my points in the EU survey response and content about it here on Lemmy https://feddit.org/post/10253134/5842417

Algorithms are the friend of profit and profit only, there's no other positive thing they are good for. People have lived without them for millenia finely, no need to continue having them.
- True. Another thing that I think is artificial about social media is anonymity. In real life you can see who somebody is when you're talking to them - you know whether they're lying about their age, or accent, or whatever. But online you could have an American pretending to be a European, or a Russian pretending to be an American, etc. And anonymity seems to encourage some people to be more abusive and insulting than they would be in real life, talking to real people.
  Anonymity might have some genuine uses though (like trying to escape persecution from your country's government).
  
  Well, seeing some politicians lie blatantly on TV with cameras and audios and the other politicans and moderators and studio guests that are physically there seeing it live, is not causing the liers to be not elected.
  These politicians are real people talking to real people all day and it does not stop them from lieing and representing inhumane and criminal policies and ideologies, while getting elected for it.
- This is an unnuanced and frankly wrong answer.
  Those people for the last several millennia didn't have immediate access to an entire globe's worth of news and information in real time... They had their local newspaper, which was weekly, maybe daily later on. Or a town crier. Being able to filter relevant stuff is important.
  
  Fair enough

I've always felt this discussion way to little mentions investigation. They call for investigation when it's about harmful chemicals in fertiliser, food, toys etc. They call for investigation when it's about fair business, corporate greed, abuse of power. Same with dangerous behaviour like drunk driving, when it's about violence from mentally unstable people. I can go on and on about times when legislators call for investigating the problem and also scientifically proving something they think is happening is actually happening but in this case it seems like all they can do is talk about examples like Cambridge analytica. Just go and prove these algorithms are doing bad things so we don't need opinion pieces in newspapers to get legislation passed. I've heard about it being 'potentially' dangerous for far too long, they could have easily gotten real answers by now.

Let's make an algorithm army, get it drunk and it will start a fight with American tech.

at this rate demsims would become recognized

This feels like a third world country demanding that the colonial power should run their schools and rise engineers so that the country can move up the production chain in a generation or two.
Edit:
If this is an online war, how could that be won with regulations? It's worse than regulating the banking sector, it's impossible to audit the constantly changing systems.
Remember when Germany made search engines pay for news? Google negotiated an agreement, all other competitors had problems.
The obvious and easiest solution would be shifiting to the fediverse. If all institutions move their accounts there would be enough momentum to make the switch.

- Jean-Marie Le Pen got 18% in 2002 apparently, but his daughter in 2022 got 41%.
  Anyway. I think it makes sense that Europe shouldn't just allow itself to be propagandised by the USA. Maybe Europe should bring in a law saying that, for example, Musk can't boost the visibility of his own account on X (here is a news story saying that he did that).
  I suppose a properly pro-democracy social media platform would be neutral and fair, rather than boosting certain types of content.

29 comments