LLMs don’t form judgments. They skip straight to the answer. No evaluation. No grounding. Just fluent output. When generation bypasses judgment, knowledge becomes a performance. Welcome to Epistemia. PNAS commentary ⬇️ www.pnas.org/doi/10.1073/...

New study on LLMs shows that while LLMs & humans converge on similar judgments of reliability of news media, they rely on very different underlying processes. In delegating, are we confusing linguistic plausibility with epistemic reliability? The age of "epistemia" www.pnas.org/doi/epdf/10....

How LLMs generate judgments www.nature.com/articles/s43... "driven by lexical and statistical associations rather than deliberative reasoning"

📢Research Highlights out today! We highlight work by @walter4c.bsky.social, @matteocinelli.bsky.social, and colleagues on how LLMs generate judgments about reliability and political bias, and how their procedures compare to human evaluation. www.nature.com/articles/s43... #cssky

Data changed the info business model: confirmation → echo chambers → infodemics LLMs drop cost of “knowledge-like” content to zero. Result: Epistemia — when language sounds like knowledge. Outsourcing shifts decisions from evidence → plausibility PNAS:https://www.pnas.org/doi/10.1073/pnas.1517441113

#Grokipedia just launched. An AI-built encyclopedia, pitched as a “neutral” alternative to Wikipedia. But neutrality is not the point. What happens underneath is. 👇

timely, considering Grokpedia and all the related implications.

Proceedings of the National Academy of Sciences@pnas.org · 11mo ago

One of the most-viewed PNAS articles in the last week is “The simulation of judgment in LLMs.” Explore the article here: https://ow.ly/7m2o50Xj6l1 For more trending articles, visit https://ow.ly/6hok50Xj6l3.

“LLMs don’t understand.” Of course. That was never the point. The point is: we’re already using them as if they do — to moderate, to classify, to prioritize, to decide. That’s not a model problem. It’s a systemic one. The shift from verification to plausibility is real. Welcome to Epistemia.

LLMs can mirror expert judgment but often rely on word patterns rather than reasoning. A new study introduces epistemia, the illusion of knowledge that occurs when surface plausibility replaces verification. In PNAS: https://ow.ly/ry7S50Xcv9b

📢Join us for a public lecture by @walter4c.bsky.social about the impacts of social media on society. ⛓️‍💥For online attendants, please register here: bit.ly/3FomgkF

1/ 🎉 First paper of 2025! In the quantitative study of the attention economy, we asked a key question: How much does a like—or a viral post—truly reverberate? Our new study, published in Scientific Reports, dives into this crucial topic. 🧵

A follow-up to this Nature paper (www.nature.com/articles/s41...) has just been published in PNAS: www.pnas.org/doi/10.1073/...

Patterns of linguistic simplification on social media platforms over time | PNAS

Understanding the impact of digital platforms on user behavior presents foundational challenges, including issues related to polarization, misinfor...

pnas.org

Breaking News! Our latest paper has just been published in PNAS! 📄 We analyzed 34 years of data a cross 8 social media platforms to uncover how language has evolved in the digital age. 👉 Read the paper here www.pnas.org/doi/10.1073/...

1/ 🚨 New preprint on arXiv 🚨 "Characterizing the Fragmentation of the Social Media Ecosystem" arxiv.org/abs/2411.16826 @matteo_cinelli , @m_starnini edoardo di martino Alessandro Galeazzi #echoplatforms #fragmentation

Hello World!