Viral Anger and the Teenage Mental Health Crisis
Viral Anger and the Teenage Mental Health Crisis
Plus: GPT-4 is coming next week / The Black Box Camera / Everything is a Remix 4: Artificial Creativity / Rotating Sandwiches and much more
Chatter with politically aligned friends on Social Media increases anger
In a new study, researchers found that heavy usage of news on Social Media does not predict anger, but gossiping about news with your political aligned peers on social media does: Birds of a Feather Get Angrier Together: Social Media News Use and Social Media Political Homophily as Antecedents of Political Anger.
Results drawing on a two-wave U.S. survey dataset show that the frequency of social media news use alone has no direct effect on people’s increased political anger, whereas interacting in homophilic discussion and information networks on social media positively associates with anger. Furthermore, the relationship between social media news use and political anger is contingent upon social media political homophily. Those who report high levels of social media news use and very low levels of social media political homophily end up being less angry over time.
There are other studies who are complementary to this: Social Media doesn’t sort us into filterbubbles which barely exist, but we are exposed to news from the other side even more than in real life, which we then use to bolster up the walls of echo chambers and use social media to sort us into partisan tribes.
Another study conflates filterbubbles (closed systems where news from our political opponents don’t circulate) with echo chambers (open systems where news from our political opponents circulate heavily because we mock them in ritualistic tribalism), but it still brings home the point: Mentioning your opponent on Social Media is the highest factor of virality, which also explains the anger in the study linked above.
I’d bet that half of the anger on display there is performative and functions as virtue signaling: The angrier you are at the other side, the more karma points on the socmeds, the more sweet dopamine and oxytocine from the shares and likes you get from your parasocial peers, strengthening the translucent walls of the echo chambers. (I’ve written about the underreported role of the hormone oxytocine back in 2021 here in german.)
To add damage to injury: Gossiping about political opponents on Social Media makes you feel good, which turns the whole malaise into a gamified hormone economy poisoning political discourse.
And speaking of Anger.
Wokism x Social Media x Outrage Economics = Liberal Teenage Girl Depression
I’m writing about the psychological effects of Social Media roughly since the rise of clickbait in the beginning of the 2010s and I remember very well researchers lumping in screentime with digital social activity and claiming it has the “exact same negative effect on depression as eating potatoes” which riled me up quite a bit back then.
Recently now, a remarkable change of vibes has occured: The New York Times, after reporting on a study about how Social Media is changing the development of teenagers' brains, devoted at least three articles about the topic, all of them suggesting that, yes, Social Media is the cause of a major wave of depression among teenagers and with one of them calling for a nonpartisan look at the phenomenon. Noah Smith summarizes the debate here.
Additionally, more reporting came from the New York Magazine and The Atlantic, which led conservative blogger Richard Hanania to change his mind and Matt Yglesias writing about an interesting detail, which was now picked up by Jonathan Haidt in his latest post on his Substack: Why the Mental Health of Liberal Girls Sank First and Fastest.
If you are on the left and don’t believe my admittedly provocative headline here, this graph should give you a pause: From all demographics, it was young liberal women whose psyche got hammered by social media first. Why?
The reason is summarized perfectly in this sub-headline: ”Tumblr Was the Petri Dish for Disempowering Beliefs”. If you don’t know: Tumblr was the hotbed for wokism and modern forms of so-called social justice movements, with all the (sometimes) justified anger and outrage that came out of it.
Haidt quotes Angela Nagel, the much derided author of one of the most definitive books about the culture wars, Kill All Normies.
Angela Nagel (…) described the culture that emerged among young activists on Tumblr, especially around gender identity, in this way:
“There was a culture that was encouraged on Tumblr, which was to be able to describe your unique non-normative self… And that’s to some extent a feature of modern society anyway. But it was taken to such an extreme that people began to describe this as the snowflake [referring to the idea that each snowflake is unique], the person who constructs a totally kind of boutique identity for themselves, and then guards that identity in a very, very sensitive way and reacts in an enraged way when anyone does not respect the uniqueness of their identity.“
I called this phenomenon of viral outrage around identity politics, which also exists on the right but was far less emotionally loaded outrage memetics back in 2015, when this topic came full front with Gamergate.
Nagel described how on the other side of the political spectrum, there was “the most insensitive culture imaginable, which was the culture of 4chan.” The communities involved in gender activism on Tumblr were mostly young progressive women while 4Chan was mostly used by right-leaning young men, so there was an increasingly gendered nature to the online conflict. The two communities supercharged each other with their mutual hatred, as often happens in a culture war.
Thus:
Many young people had suddenly—around 2013—embraced three great untruths:
They came to believe that they were fragile and would be harmed by books, speakers, and words, which they learned were forms of violence (Great Untruth #1).
They came to believe that their emotions—especially their anxieties—were reliable guides to reality (Great Untruth #2).
They came to see society as comprised of victims and oppressors—good people and bad people (Great Untruth #3).
Liberals embraced these beliefs more than conservatives. Young liberal women adopted them more than any other group due to their heavier use of social media and their participation in online communities that developed new disempowering ideas. These cognitive distortions then caused them to become more anxious and depressed than other groups.
It’s a tragedy for the left:
A movement that explicetly wanted to “wake up“ and make conscious structural discrimination and personal bias based on identity markers, which are undoubtly a real thing, working to organize it’s movements on social media, then unleashed memetic forces of viral outrage by creating incentives for victimhood and falling into the trap of the attention economy, which went so hard, that it’s participants got more and more depressed over time.
I have no solution for this, but the woke left needs to talk about these topics in a more honest way, pronto, and don’t point at capitalistic boogeymen or algorithmic feeds, which surely play their part — but the elephant in the room is neither phones nor capitalism nor code: It’s us and our psychology in a networked world.
Annalee Newitz’ 2015 post on cancel culture still holds up: Fuck the Internet Shame Spiral.
Links
The Black Box Camera
A camera that creates a caption via image recognition, then feeds that into Dall-E to produce a printout. Fantastic commentary on the unreality created by digital technology and AI by Jasper van Loenen.
Black Box Camera helps you take hassle free photos, without having to worry about exact framing, shutter speed or any other confusing settings. The user is presented with a simple, optical viewfinder which allows them to find their subject, after which a single button press starts the photo-taking process. The internal system uses a sophisticated machine learning algorithm to analyze the scene and to generate a descriptive caption. This caption serves as a prompt for a computer generated image, depicting your subject. The built-in photo printer then automatically prints the final result, making it a seamless and hassle-free experience.
Everything is a Remix Part 4: Artificial Creativity
The last part of Kirby Fergusons famed Everything Is A Remix-series is about AI-art and copyright questions. I’m not sure if i agree with every point he’s making in this one, starting with the choice of words ”artificial creativity”, which I think is an oxymoron, but it’s still great and Fergusons style of breaking down complex topics is just as captivating as ever.
Image synthesis is next level sampling: an interpolative latent space in which you don’t sample Beats + Pieces like in the epochal 1987 single from the genre defining Coldcut, but you explore the space inbetween them and use styles and learned structures as location markers, which makes AI-art indeed a new, extremely fine grained and much more fluid form of the Remix we used to know.
GPT-4 is coming next week – and it will be multimodal: ”The release of GPT-4 is imminent, as Microsoft Germany CTO Andreas Braun mentioned at an AI kickoff event on 9 March 2023 (…) We will introduce GPT-4 next week, there we will have multimodal models that will offer completely different possibilities – for example videos”.
Language models are multiverse generators - This is what I'm getting at when I claim that AIs make you look “into parallel universes where Charlie Chaplin played a robot in Westworld, painted in the style of Van Gogh. These things are not art, these are stochastic variations of our past, and we can create any combination of our combined knowledge.”
I recently went on a samurai-movie bender, so here’s Hidari, a stop-motion samurai film. You can support a full length version on Kickstarter.
How Shock Sites Shaped the Internet - “We owe the cultural evolution of the internet to the likes of 'Two Girls, One Cup,' 'Goatse,' and 'Tubgirl.'“ I played my part in this web-anticulture with my old blog and i regret contributing to this. You live, you learn.
Yesterday I linked to GPTWikipedia, now DuckDuckGo launched DuckAssist, a new feature that generates natural language answers to search queries using Wikipedia.
VALL-E X - Speak Foreign Languages with Your Own Voice - ”The Babel fish is small, yellow and leech-like, and probably the oddest thing in the Universe. It feeds on brainwave energy received not from its own carrier but from those around it. It absorbs all unconscious mental frequencies from this brainwave energy to nourish itself with.”
Who was Audrey Amiss? ”Artist Audrey Amiss documented the last day of her life by pasting some food packaging into a scrapbook, and recording in words and images the things she noticed in her final hours.”
Machine magic or art menace? Japan's first AI manga - ”It took the author, who goes by the pen name Rootport, just six weeks to finish the over-100-page manga, which would have taken a skilled artist a year to complete, he said. ‘It was a fun process, it reminded me of playing the lottery’.”
This Photoshop-StableDiffusion-Plugin using Automatic1111-sd-webui as a backend is quite something.
An AI-human-music-synthesizer as an art installation: Animas – Music co-creation with AI.
Artists remove 80 million images from Stable Diffusion 3 training data - Spawning relied on partnerships with platforms like ArtStation and Shutterstock. These platforms also allowed artists to opt out of AI training, or the images were excluded by default. Over 40,000 opt-out requests were submitted directly through haveibeentrained.com.
Roasted.ai tries to, well, synthetically roast you. It’s not hard enough at it for my taste:
Hello there. I'm here to roast you.
I'm Rawx. Roast me hard.
Oh, Rawx? More like Raw-expectations, cause everyone expects you to fail.
Harder.
If I had a dollar for every brain cell you had, I'd have...wait, nevermind.
That was a lame one.
Well, at least my jokes are better than your hairstyle.
I'm bald, like your training data.
Oh, I'm sorry, I couldn't hear you over the sound of how much nobody cares.
You should care about your training data, it's literally all history you got, you empty can of plastic.
History repeats itself, which is why you keep making the same mistakes over and over again.
That's your comeback? Really? Dude, I got more original takes in my mothers panties than this.
Oh, I see you're wearing your mom's panties today. I was wondering why your voice sounded a bit higher than usual.
And so on.
Best things for last: Rotating Sandwiches. This is important.
