PhD student in @nerdsitu.bsky.social @itu.dk. 🔮
Latest work on the impact of generative AI on social media, an experimental study: https://ai-research.andersgiovanni.com/
Key Finding: Generative AI broadens participation and facilitates user interaction, but risks saturating platforms with perceived lower-quality, generic content, leading to diminishing authenticity, trust, and diversity of perspectives.
Check out the explainer: ai-research.andersgiovanni.com
No fewer than three people were needed to cover all the aspects of our dialogue simulation paper. Thanks for the interest — check out the preprint. Link in Dustin’s post.
@dustinbwright.com@ic2s2.bsky.social#ic2s2
We had a great time talking about dialogue simulation with LLMs at @ic2s2.bsky.social !!! Amazing work by all of our colleagues at UMich.
See the preprint of this work here: arxiv.org/abs/2409.08330
What actually happens on social media when users get access to integrated AI?🤖
In our new preprint, we present a controlled experiment testing the impact of generative AI (4 treatments + control) on a social media platform, using a representative sample of nearly 700 U.S. participants.
It’s a wrap ✨
One more edition of #icwsm is over. Amazing research outputs, insightful conversations and many new, exciting ideas!
Thanks to all the organizers, chairs and volunteers who made the event possible, and to all the participants for making ICWSM 2025 a memorable edition 🇩🇰
Excellent story in @theverge.com today on the impact of AI-Mediated Communication. Quotes me but the best line is the closer from the writer, Sara Parker.
Testing LLMs in the Prisoner’s Dilemma to reveal their moral biases. LLMs can defect often but become more forgiving and less retaliatory than humans when opponents are mostly cooperative. Nicolò Fontana at #ICWSM. w @frapierri.bsky.social
Paper: ojs.aaai.org/index.php/IC...
Fresh evidence agreeing w/ one of our studies mentioned in the story above, from @danielmromero.bsky.social@davidjurgens.bsky.social@lajello.bsky.social
"AI-tools increase user engagement and volume of generated content, but... decrease the perceived quality and authenticity of discussion"
AI-assisted social media posting increases the volume of content people produce. However, that content is perceived as lower quality than purely human-generated posts. A dangerous spiral that may flood social media with low-quality content. Results from a controlled study: arxiv.org/abs/2506.14295
What actually happens on social media when users get access to integrated AI?🤖
In our new preprint, we present a controlled experiment testing the impact of generative AI (4 treatments + control) on a social media platform, using a representative sample of nearly 700 U.S. participants.
I'll be at #ICWSM 2025 next week to present our paper about Bluesky Starter Packs.
For the occasion, I've created a Starter Pack with all the organizers, speakers, and authors of this year I could find on Bluesky!
Link: go.bsky.app/GDkQ3y7
Let me know if I missed anyone!
What actually happens on social media when users get access to integrated AI?🤖
In our new preprint, we present a controlled experiment testing the impact of generative AI (4 treatments + control) on a social media platform, using a representative sample of nearly 700 U.S. participants.
We find a clear duality between the use of AI tools and participants’ perception of the content. All that and much more on our blog and in the paper:
🧠 Blog: ai-research.andersgiovanni.com
📄 Arxiv: arxiv.org/abs/2506.14295
"Urban Highways Are Barriers to Social Ties" out on PNAS!
The 1st large-scale measure of how highways weaken social connections between the communities they separate. This barrier effect is strong in the 50 largest US cities--especially for low-income Black communities.
www.pnas.org/doi/10.1073/...
The second edition of my Atlas for the Aspiring Network Scientist is out! You should check it out if you like to learn about network science or graph neural networks. Check out the blog post I wrote to present it: https://www.michelecoscia.com/?p=2393
🚨 Spread the word! 🚨 The #ICWSM Adamic-Glance Distinguished Young Researcher Award is now accepting nominations and self-nominations! 🌟
📅 Deadline for this year's award: March 15th
More info: www.icwsm.org/2025/submit/...
📣 The PhD Open Call 2025 is now open! Get in touch with us to apply for a PhD with NERDS in beautiful Copenhagen. Deadline: Feb 24th
nerds.itu.dk/2025/01/31/p...
🔦 Researcher Spotlight: Arianna Pera
👋 Meet Arianna, PhD Student in NERDS. Arianna has been in the group for 2 years.
🌟 What does Arianna do: Studying how collective action can arise from social media interactions.
Meet NERDS 👋 We're 27 researchers at ITU Copenhagen studying how data, networks, and AI shape society!
We use data science to decode everything from social networks to urban systems - and we're big fans of bikes! 🚲
Stay tuned as we share more about the cool stuff we're working on! nerds.itu.dk
Happy 2025! 🎉
📅 The ICWSM submission deadline is just two weeks away (Jan 15)!
Check the page limit here: www.icwsm.org/2025/submit/...
Don't forget the mandatory Ethics Checklist and that optional Appendices don’t count toward the limit
My TEDx talk about people-centered AI and mental health is finally live! This was a really big accomplishment with a lot of work, and I'm proud of it, esp in light of conversations on social media and mental health we're having today - www.youtube.com/watch?v=H30b...
What a nice blogpost from Hugging Face (@hf.co) on test-time "thinking" with open-source models, showing how a 1B model can beat an 8B and how a 3B can beat a 70B with the right strategies. It also gives a great overview of different approaches - check it out! 🚀
huggingface.co/spaces/Huggi...
Not a fan of this notion of models “thinking,” but it does make the concept of adding intermediate steps before delivering the output much easier to grasp. It seems like an amazing way to achieve more with less! 👏🏼
📢 New preprint: #network analysis of the #danish#Bicycle infrastructure 🚲🕸️ -Yes, a *whole country's* bike network analyzed, led by @ane. See the 37-page paper: https://arxiv.org/abs/2412.06083
Main results: Good bikeability in cities, some regional […]
[Original post on datasci.social]
The Call for Papers is live on our website! 🎉 We’re accepting submissions for full papers, demos, posters, and datasets—all due by January 15. Don’t miss the chance to showcase your work and join us this June in beautiful Copenhagen 🇩🇰
www.icwsm.org/2025/submit/...
Quick tip: use uv to handle your envs, packages, and projects. I've used poetry for quite some time now, it's just pretty convenient even though it is slow.
uv is, as they say on their website, "An extremely fast Python package and project manager" - this is true. 💨
You're welcome!👏🏼
Podcasts are a popular medium, but data for computational research is limited! We introduce the Structured Podcast Research Corpus (SPoRC - huggingface.co/datasets/bli...), a large, multimodal dataset of English podcasts 🧵
arxiv.org/abs/2411.07892
It's great to see the massive wave of people coming to 🦋.
Here's my first stab at a Computational Social Science starter pack:
go.bsky.app/2m6iQ68
Go check out similar lists to bootstrap your network:
bsky.app/starter-pack...bsky.app/starter-pack...
Wonder if these lists could be merged somehow..
NERDS leave X, join Bsky and LinkedIn
nerds.itu.dk/2024/11/12/n...
"The media’s the most powerful entity on earth. They have the power to make the innocent guilty and to make the guilty innocent, and that’s power. Because they control the minds of the masses."
– Malcolm X (formerly Malcolm Twitter)