pettertornberg.com
Petter Törnberg
@pettertornberg.com
4.4K followers 160 following 400 posts
Assistant Professor in Computational Social Science at University of Amsterdam Studying the intersection of AI, social media, and politics. Polarization, misinformation, radicalization, digital platforms, social complexity.
Posts Media Videos Starter Packs
Pinned
pettertornberg.com
Petter Törnberg @pettertornberg.com · Jan 14
Misinformation isn't random - it's strategic. 🧵

In the first cross-national comparative study, we examine 32M tweets from politicians.

We find that misinformation is not a general condition: it is driven by populist radical right parties.

with @julianachueri.bsky.social
doi.org/10.1177/1940...
160 3.4K 6.4K
pettertornberg.com
Petter Törnberg @pettertornberg.com · 2d
Wait, Trump eats the pizza from the crust inward?! 😧
pettertornberg.com
Petter Törnberg @pettertornberg.com · 8d
@jerusalem.bsky.social Will The Argument bring back 'Good on Paper' in some form? Please do, its loss has left a huge hole in my podcast app!

And there are a lot of persuasive papers out there!
1 1
Reposted by Petter Törnberg
veganfuturist.bsky.social
Nikola Danaylov Ⓥ @veganfuturist.bsky.social · 9d
🚨 New #SingularityFM 🎙

What do Facebook, Google & TikTok really see when they look at us — and what do they miss?

My latest interview with @pettertornberg.com explores algorithmic tyranny, digital modernity & the future of power.

👉 snglrty.co/4mDn1Gb
Petter Törnberg: Algorithmic Tyranny & Digital Modernity
See this interview to discover Petter Törnberg’s insights on algorithmic tyranny, digital modernity, and how platforms reshape power.
snglrty.co
1 1
Reposted by Petter Törnberg
rawx.bsky.social
René Walter @rawx.bsky.social · 10d
Not just one but two podcasts featuring @pettertornberg.com who wrote some of the most interesting papers in recent years.

1: @seanmcarroll.bsky.social, below
2: Singularity Weblog: Algorithmic Tyranny, the Rise of Digital Modernity and Seeing Like a Platform www.youtube.com/watch?v=lJrV...
1 4
pettertornberg.com
Petter Törnberg @pettertornberg.com · 10d
Such a crazy and amazing experience to be on a podcast of which I'm a huge fan!
seanmcarroll.bsky.social
Sean Carroll @seanmcarroll.bsky.social · 10d
Mindscape 330 | Petter Törnberg @pettertornberg.com on the Dynamics of (Mis)Information. #MindscapePodcast

www.preposterousuniverse.com/podcast/2025...
Title card for Mindscape podcast episode with Petter Törnberg
3 7 45
Reposted by Petter Törnberg
julianachueri.bsky.social
Juliana Chueri @julianachueri.bsky.social · 12d
www.flamman.se/valfardsstat...

It was great to talk to @flamman.se about the link between weak welfare state protection and the expansion of online platform work in Europe. @pettertornberg.com
Välfärdsstat ut, gigjobb in
Gigjobb och nedmonterad välfärd går hand i hand
www.flamman.se
2 8 17
pettertornberg.com
Petter Törnberg @pettertornberg.com · 20d
No.
I said I obviously wouldn’t review a revised version of a paper I’d flagged as AI.
I just received a reply from the journal
asking for a suggestion for an alternative reviewer. 😒
1
pettertornberg.com
Petter Törnberg @pettertornberg.com · 20d
And I now received an automatic request to review a revised version.

This is depressing. Academia is becoming AI slop.
pettertornberg.com
Petter Törnberg @pettertornberg.com · Sep 1
I flagged to the editor that the paper I reviewed looked AI-generated.

Immediately received a generic AI-generated email from the editor.

The AI future is here!
4 1 5
pettertornberg.com
Petter Törnberg @pettertornberg.com · 22d
Yes, as long as the bots also vote in the US election, they will be included in the ANES data! :)

But kidding aside, the posts you see on 'X' are indeed likely to be even more right-wing leaning - due to documented algorithmic biases etc.
2
Reposted by Petter Törnberg
environmentalpol.bsky.social
Environmental Politics @environmentalpol.bsky.social · 22d
New article!

The aesthetics of climate misinformation: computational multimodal framing analysis with BERTopic and CLIP by Anton Törnberg & Petter Törnberg / @pettertornberg.com

doi.org/10.1080/0964...
ABSTRACT Climate misinformation today is increasingly conveyed through multimodal content, where images and text interact to shape meaning, emotional resonance, and enhance credibility. Yet, research on climate misinformation has rarely explored how these modes work in tandem. This paper introduces a multimodal analytical framework to examine climate-related visual misinformation. Analyzing 17,848 image-text posts, we combine BERTopic, CLIP, and qualitative framing analysis to investigate how climate denialist narratives are constructed. Our findings reveal a paradoxical communication strategy: the movement appropriates scientific aesthetics – graphs, statistics, and technical imagery – to contest the scientific consensus, projecting rationality, authority, and masculine self-control. In contrast, climate advocates are depicted through emotionally charged, feminized, and irrational imagery. These contrasting multimodal framings supports a strategy of strategic depoliticization, presenting ideological claims as neutral, objective truths. We argue that understanding climate misinformation requires moving beyond factual accuracy to examine the multimodal forms through which it gains legitimacy and emotional power.
5 23
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Not really possible to see in the data unfortunately!
1
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
The one-plot summary of what Elon Musk did to Twitter.
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
This is where it gets wild 👇

In 2020, Twitter use was highest among people who loved Democrats and disliked Republicans.

By 2024, it completely flipped: the more polarized Republican you are, the more you use Twitter/X.

From blue stronghold → red megaphone. All for just $44 billion.
1 21 49
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Curious to see other breakdowns? Let me know!

I’ve got the ANES data open and I need your help to keep procrastinating — I have so much important work to avoid doing! 📊
1 4
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Who posts about politics on Facebook?

🔍 More affectively polarized voters post more - making our politics seem more polarized than it is (the infamous 'social media prism' @chrisbail.bsky.social)

📉 By 2024, political posting on Facebook dropped across the board — but especially among Democrats.
1 2
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Looking closer at the frequency of use, Facebook shifted right too.

From 2020 → 2024:
🔵 Strong Democrats use it less
🔴 Strong Republicans use it more

The platform’s core audience is now clearly more Republican.
1 4
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Overall platform reach, 2020 → 2024 👇

📱 Facebook & YouTube: still near-universal, little change.
📱 Instagram & TikTok: surged, especially among Dems.
📱 Twitter/X: flat or declining, biggest drop among Dems.
📱 Reddit: grew a lot on the left.

The partisan sorting shows up in who sticks around.
1 1 7
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Here’s the partisan lean of platform audiences, 2020 → 2024.

📊 Instagram, TikTok, Reddit stay solidly Dem.
📊 Facebook & YouTube: modestly rightward.
📊 Twitter/X: sharpest move right.
📊 “Other” (Threads, Bluesky, etc.): flipped hard from red to blue.

The sorting is unmistakable.
1 3 13
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Looking at party affiliation, the partisan map hasn’t moved much outside of Twitter/X:

🔵 Instagram, TikTok, Reddit → firmly Democratic
⚪ Facebook, YouTube → still mixed
🔴 Twitter/X → the big flip

The social media ecosystem has become more sorted.
1 4
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
Posting on Twitter tells the same story.

🔵 In 2020, posting on Twitter was correlated with being polarized Democrat.
🔴 By 2024, that gap reversed — Democrats pulled back, Republicans leaned in.

Political posting on Twitter/X is now correlated with being polarized Republican.
1 1 6
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
This is where it gets wild 👇

In 2020, Twitter use was highest among people who loved Democrats and disliked Republicans.

By 2024, it completely flipped: the more polarized Republican you are, the more you use Twitter/X.

From blue stronghold → red megaphone. All for just $44 billion.
1 31 60
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
But Snapchat and TikTok also saw significant movements rightward between 2020 → 2024:

They're still Democratic-leaning overall, but the change is clear: the user base has become more Republicans.
2 1 4
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
🚨 The Democratic exodus from Twitter/X is real.

By 2024, Twitter/X had flipped to a Republican-leaning platform.

Where did Democrats go? "Other" platforms — Threads, Bluesky, Mastodon — which shifted from majority Republican (~62%) in 2020 to majority Democratic (~59%) in 2024.
2 6
pettertornberg.com
Petter Törnberg @pettertornberg.com · 23d
How much did Elon's takeover reshape Twitter/X? How did the partisan tilt of social media use change from 2020 to 2024?

The ANES 2024 data is out — and this thread answers all your burning questions! 🔥
1 29 42
pettertornberg.com
Petter Törnberg @pettertornberg.com · 24d
But less issue as here, as we focus on structural features, like network properties.
Also, it's actually really hard to get an LLM to get into "extremist poster subspace"; the real problem is that they're too polite.

Working on validating with conventional ABMs + human participant experiments.
1
pettertornberg.com
Petter Törnberg @pettertornberg.com · 24d
I actually largely agree about the validity challenges with GABM.

We even wrote a paper on this precise issue: arxiv.org/abs/2504.03274
Do Large Language Models Solve the Problems of Agent-Based Modeling? A Critical Review of Generative Social Simulations
Recent advancements in AI have reinvigorated Agent-Based Models (ABMs), as the integration of Large Language Models (LLMs) has led to the emergence of ``generative ABMs'' as a novel approach to simula...
arxiv.org
1 1