Lightnews — Scholar-powered news

Reposted by Angel Hsing-Chi Hwang

Alex Hanna-ween @alexhanna.bsky.social · Aug 4

I would also like to remind folks that OpenAI wrote a paper in which they prompted GPT-4 on which jobs they thought would be most exposed to automation.

They validated it by comparing it to responses that people who worked OpenAI gave to the same question.

arxiv.org/abs/2303.10130

6 49 130

Reposted by Angel Hsing-Chi Hwang

Aaron Roth @aaroth.bsky.social · May 17

Check out the terrific set of EC 2025 accepted papers! ec25.sigecom.org/program/acce...

EC 2025 Accepted Papers - EC 2025

1. Optimality of Non-Adaptive Algorithms in Online Submodular Welfare Maximization with Stochastic Outcomes Authors: Rajan Udwani (University of California, Berkeley) 2. Investment and misallocation i...

ec25.sigecom.org

11 22

Reposted by Angel Hsing-Chi Hwang

Columbia Journalism Review @columjournreview.bsky.social · May 12

AI is already changing how journalists operate. Reporters, editors, executives, and others across the news industry share their advice on how to engage—and where to draw the line. By @mikeananny.bsky.social
and @mattdpearce.com with USC's AI for Media & Storytelling. www.cjr.org/feature-2/ho...

How We’re Using AI

The rapid development of AI is already changing how journalists operate. Reporters, editors, executives, and others across the news industry share their advice on how to engage—and where to draw the l...

www.cjr.org

3 7

Reposted by Angel Hsing-Chi Hwang

John Holbein @johnholbein1.bsky.social · May 1

If you run conjoint experiments, you need to read this.

Most conjoints estimate average effects for each attribute.

But what if the effect of one attribute depends on the others?

This paper has got you covered!

4 11 51

Reposted by Angel Hsing-Chi Hwang

Mike Ananny @mikeananny.bsky.social · Apr 24

So much is broken right now, but I want to share an amazing new set of short, teachable interdisciplinary essays on

** Sociotechnical Error **

Live at IJOC journal @ijoc-usc.bsky.social: ijoc.org/index.php/ij... (scroll to Forum)

Intro by me & Simogne Hudson: ijoc.org/index.php/ij...

Pls share!

Headline: " International Journal of Communication Publishes a Forum on "Oops? Interdisciplinary Stories of Sociotechnical Error""

Abstract: " What can we learn about people and technology through interdisciplinary stories of sociotechnical errors, failures, breakdowns, and mistakes?

Guest edited by Mike Ananny and Simogne Hudson, the Forum on Oops? Interdisciplinary Stories of Sociotechnical Error takes up the question through a playful and provocative mix of projects that show how sociotechnical errors happen, why they matter, and what they reveal about people, technology, and power. Amidst so many complex collisions among people, data, engineering, and media—and in an age when technological "innovation" is widely celebrated and inescapable—these articles offer changes to pause and ask what system failures show about how people and machines intersect and vie for power.

Including scholars from communication, media studies, urban planning, critical data studies, and science and technology studies, the collection of essays invites readers to see failures anew—to consider errors, breakdowns, and mistakes from a different perspective, method, or normative stake. Use these essays to start conversations about what "error" means in your work or community, and why it matters.

We invite you to read these articles that published in the International Journal of Communication on April 23, 2025. Please log into ijoc.org to read the papers of interest. We look forward to your feedback!"

List of authors and essay titles:

Oops? Sociotechnical Errors as Interdisciplinary Stories of Complex Relations, Shared

Consequences, and Resilient Hopes—Introduction

Mike Ananny, Simogne Hudson

Uncertainty as Spectacle: Real-Time Algorithmic Techniques on the Live Music Stage
Stephen Yang

When Faulty AI Falls Into the Wrong Hands: The Risks of Erroneous AI-Driven Healthcare Decisions

Eugene Jang

Fake It Till You Make It: Synthetic Data and Algorithmic Bias

Sook-Lin Toh, Jiwon Park

Discourses of Sociotechnical Error and Accuracy in U.S. and PRC News Media: The Case of the 1999
Bombing of the Chinese Embassy in Belgrade

Max Berwald

Affective Experiences of Error

Megan Finn, Youngrim Kim, Ryan Ellis, Amelia Acker, Bidisha Chaudhuri,

Stacey Wedlake

Peeling Back the Layers of “Paint on Rotten Wood”: Unraveling the Senate’s “Big Tech and Child

Sexual Exploitation Crisis” Hearing

Kyooeun Jang

Kicking Error Out of the Game: Video Assistant Referee as Technosolutionism

Pratik Nyaupane, Alejandro Alvarado Rojas

When User Consent Fails: How Platforms Undermine Data Governance

Rohan Grover

Ephemeral Platforms, Enduring Memories: Errors and Digital Afterlife

Sui Wang

:Chatting: Errors in Live Streamer Discord Servers

Kirsten Crowe

Hole in the (Pay)Wall: Monetized Access, Content Leaks, and Community Responsibility

Celeste Oon

Edges, Seams, and Ecotones: Error in Interstate Landscapes

Cindy Lin, Steve J. Jackson

Quantifying Housing Need in California: The Erroneous Practice of Evidence-Based Policy

Elana R. Simon

1 8 17

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

Featuring all-⭐️ panelists✨ @mbernst.bsky.social, Shyam, Renwen, @manoelhortaribeiro.bsky.social, Yingdan, @serinachang5.bsky.social @manoelhortaribeiro.bsky.social @sherrytswu.bsky.social Aimei, @joon-s-pk.bsky.social Dmitri @ognyanova.bsky.social @ziangxiao.bsky.social Ayman @aaronshaw.bsky.social

3

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

3️⃣ How can researchers approach addressing homogeneity, biases, and ethical concerns of LLM simulation output?

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

2️⃣ Whether/how can researchers scale insights of LLM simulations of individuals' responses to study group and even network patterns?

1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

1️⃣ When/how can researchers integrate the use of LLM simulation and synthetic data into existing human subjects research pipelines? How do we perform evaluation accordingly?

1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

This panel will discuss the opportunities and perils regarding the use of LLM, simulation, and synthetic data for human subjects research. We will break the discussion down into three themes/challenges:

1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 22

📣 Calling all #CHI2025 attendees who work with human participants: Join our panel discussion on #LLM, #simulation, #syntheticdata, and the future of human subjects research on Apr 30 (Wed), 2:10 - 3:40 PM (JP Time)

Post your questions for panelists here: forms.gle/m2mXY3xFafAX...

2 5 18

Reposted by Angel Hsing-Chi Hwang

Data & Society @datasociety.bsky.social · Apr 11

May 6, in NYC or online: Join @posada.website and Labor Futures Program Director @aihathing.bsky.social as they discuss the uneven effects of AI technologies across industries and on a broad diversity of workers. Learn more and RSVP! datasociety.net/events/what-...

A yellow promotional graphic for the event “What is Work Worth” happening on May 6 at 5pm ET in NYC and on Zoom, with Dr. Julián Posada and Aiha Nguyen.

2 5

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 19

Would love to stop by if time permits!

1 1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 19

Thanks for sharing our work, Freddy!!

1 1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 18

Additionally, @dohyojin.bsky.social, Jessica He,
@feldmanmolly.bsky.social, Seyun Kim, and I are organizing a workshop at #CHIWORK on "Navigating Generative AI Disclosure, Ownership, and Accountability." Check out more info here (chiwork-aidisclosure.github.io), and we would love to see you there!!

Navigating Generative AI Disclosure, Ownership, and Accountability in Co-Creative Domains

chiwork-aidisclosure.github.io

5

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 18

To further develop this workstream, I will present our latest findings and seek feedback at
#AOM, @ic2s2.bsky.social, and @datasociety.bsky.social 's upcoming workshop on "What is work worth?" See extended abstract here: angelhwang.github.io/doc/ic2s2_AI...

angelhwang.github.io

1 5

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 18

Yao-Yuan Yang and I verified this concern by tracking the performance of 9,149 freelancers across two platforms (Upwork and Bēhance): Creators who declare the use of AI receive significantly lower pay, but non-creatives jobs earn more by labeling themselves as "AI Pros."

1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 18

This ongoing work is inspired by my favorite project with @qveraliao.bsky.social, Su Lin Blodgett, @aolteanu.bsky.social, and Adam Trischler. Writers felt they could preserve their authentic voice but worried audiences would not value AI-assisted work as much as solo work. arxiv.org/abs/2411.13032

"It was 80% me, 20% AI": Seeking Authenticity in Co-Writing with Large Language Models

Given the rising proliferation and diversity of AI writing assistance tools, especially those powered by large language models (LLMs), both writers and readers may have concerns about the impact of th...

arxiv.org

1

Angel Hsing-Chi Hwang @angelhwang.bsky.social · Apr 18

Starting my journey on Bluesky with a topic that I care deeply about: AI tools can support creators in various ways, but disclosing AI use may risk devaluing creative work.

Check out our abstract here: angelhwang.github.io/doc/ic2s2_AI...
Inspired by our past work: arxiv.org/abs/2411.13032

"It was 80% me, 20% AI": Seeking Authenticity in Co-Writing with Large Language Models

Given the rising proliferation and diversity of AI writing assistance tools, especially those powered by large language models (LLMs), both writers and readers may have concerns about the impact of th...

arxiv.org

1 5 25