Markets typically give consumers what they want, but the news market is different, with a lot of dissatisfaction. Can we solve some of these frictions with news customization? To find out, we built an AI-powered news app that allows users to customize the news. Details below! 1/5 The app unbundles presentation from coverage. Users can make the news more fact-driven and less opinionated ("Facts Only") or make it more slanted (conservative or liberal). Or they can make it less complex, more entertaining, or change the sentiment (positive or negative). 2/5

OpenAI has launched Operator, an agent that can perform tasks in your browser. I asked it to complete a Qualtrics survey I created. The results are very promising for Operator but *very* concerning for survey researchers On the first question, I included a standard attention check. It had no problem picking the top and bottom options. It didn't even bother asking me for help on this.

Some LaTeX tips for beginners (and not-so-beginners) Create a file called preamble.tex to store all your packages and settings and keep the main file tidy. In main.tex, add \input{preamble.tex}

Ph.D. students in economics: How to organize your projects and not make the replication files a huge pain - a thread with a minimal working example with some potential extensions. First, create four folders: one for raw data (csv files), one for code (your do files), one for data (your cleaned files), and one for documentation (e.g. qsv files for Qualtrics)

1/ People often say that null results are penalized in the publication process. Is that true? In a new paper with @cp_roth, @FelixChopra, and Andreas Stegmann, we examine whether and why that's the case! Read on to learn more! 2/ We recruit a sample of more than 500 economists and ask them to evaluate different hypothetical research studies. We vary whether a given study had a large and statistically significant main effect or a low and not statistically significant main effect.