ef110econ.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

Recently accepted by #QJE, “Generative AI at Work,” by Brynjolfsson (@erikbryn.bsky.social), Li, and Raymond (@lindseyraymond.bsky.social): doi.org/10.1093/qje/...

submitted 6 days ago • 0 comments

Forget “tapestry” or “delve” these are the actual unique giveaway words for each model, relative to each other. arxiv.org/pdf/2502.12150

submitted 8 days ago • 7 comments

Last year, we published a paper showing that AI models can "debunk" conspiracy theories via personalized conversations. That paper raised a major question: WHY are the human<>AI convos so effective? In a new working paper, we have some answers. TLDR: facts osf.io/preprints/ps...

submitted 8 days ago • 19 comments

What I mean with concrete solutions that should be tested and developed at scale: at the Prosocial Design Network we keep on collecting scientific evidence for what design elements of digital platforms promote prosocial behaviour, see the library of interventions: www.prosocialdesign.org

submitted 9 days ago • 1 comment

There is a lot of important stuff in this new paper by Anthropic that shows how people are actually using Claude. 1) The tasks that people are asking AI to do are some of the highest-value (& often intellectually challenging) 2) Adoption is uneven, but many highly compensated fields already are high

submitted 15 days ago • 2 comments

Looking forward to reading this study in full, but have a feeling the headline is likely over claiming. Not that this outcome is not possible on some scale but from what I read in the 404 summary, a self report survey is not reliable enough to make this kind of claim.

submitted 16 days ago • 0 comments

Hat tip to @acjsissons.bsky.social for posting this into my feed. But in my view the performance gap of heat pumps, insulation and other green tech is not a niche issue, or at least shouldn't be! This report shows performance gaps can be as big as 50%!!! Other papers show more! 1/2

submitted 20 days ago • 1 comment

Try it yourself! regcheck.app

submitted 20 days ago • 0 comments

NEW @nature.com article w/Abhijit Banerjee, Esther Duflo, Liz Spelke + fantastic team! ☑️Children working can use arithmetic to solve problems in their shops, but struggle with abstract math ☑️Children in school can do abstract math, but cannot apply it www.nature.com/articles/s41...

submitted 20 days ago • 9 comments

Are you running AB testing studies on Facebook or Google? Or reviewing papers using them? Check out our open access paper, On the Persistent Mischaracterization of Google and Facebook A/B Tests: How to Conduct and Report Online Platform Studies at doi.org/10.1016/j.ij... @boegershausen.bsky.social

submitted 28 days ago • 0 comments

This work on behavioral systems mapping looks great! Adding this to my reading list for the blog post on I frame vs s frame I am working on. Thinking a lot about the role of formal and informal power in these systems and methods researchers can use to measure it.

submitted 28 days ago • 1 comment

After ending Meta's fact-checking partnerships, Mark Zuckerburg said that it had never been broadly accepted. According to a new poll by @bostonu.bsky.social, nearly 2 in 3 US adults agree that fact-checkers should verify claims on social media. A 🧵 (1/n). www.bu.edu/articles/202...

submitted 29 days ago • 8 comments

People often claim they know when ChatGPT wrote something, but are they as accurate as they think? Turns out that while general population is unreliable, those who frequently use ChatGPT for writing tasks can spot even "humanized" AI-generated text with near-perfect accuracy 🎯

submitted 29 days ago • 9 comments

🚨OpEd+data: Meta is out of step with public opinion🚨 Zuck cut moderation b/c he said people no longer want it. But he's wrong! We polled 1k Americans and most people, including majority of Reps: i) want content moderation ii) don't want Community Notes w/o fact-checkers thehill.com/opinion/tech...

submitted 29 days ago • 6 comments

Next big thing for brands: knowing what sites agents prefer. If you ask for stock prices, Claude with Computer Use goes to Yahoo Finance while Operator does a Bing search Operator loves buying from the top search result on Bing. Claude has direct preferences like 1-800-Flowers We don't know why

submitted 34 days ago • 7 comments

Our global study on the state of trust in scientists is now out in Nature Human Behaviour! 🥳 With a team of 241 researchers, we surveyed 71,922 people in 68 countries, providing the largest dataset on trust in scientists post-pandemic 👇🧵https://www.nature.com/articles/s41562-024-02090-5

submitted 37 days ago • 24 comments

Paper alert 🚨 Our experiment on LLM search, designed before we knew Microsoft or Google would go into LLM search, was just accepted at CHI arxiv.org/abs/2307.03744

submitted 41 days ago • 0 comments

Big shift 2️⃣ for behavioural science: make it more social @mhallsworth.bsky.social shares how examining social connections, values, and identities can lead to surprising insights and push the field forward. Read the second in our 4-part series👇 bit.ly/4jlRsR8

submitted 41 days ago • 0 comments

New randomized, controlled trial by the World Bank of students using GPT-4 as a tutor in Nigeria. Six weeks of after-school AI tutoring = 2 years of typical learning gains, outperforming 80% of other educational interventions. And it helped all students, especially girls who were initially behind.

submitted 42 days ago • 15 comments

Interested in #trust in #AI? 🤖 Check out this article about our recent publication in JEP:G! #trustresearch 👇

submitted 48 days ago • 0 comments

What is one invention that should never have come to fruition because there's no way it could have ever passed even the most basic cost benefit analysis? I'll go first: Leaf blowers.

submitted 51 days ago • 0 comments

What happens to US firms barred at random from hiring immigrants? Delighted that my research w/ Ethan Lewis is part of the #ASSA2025 immigration policy this weekend Our summary at @voxeu.org—> cepr.org/voxeu/column... Also new work from @kevinshih.bsky.social & Chen et al. cc @piie.com #EconSky

submitted 55 days ago • 2 comments

2024 marks the official launch of the Human-AI Collaboration Lab, so I wrote a one page letter to introduce the lab, share highlights, and begin a lab tradition of reflecting on the year and sharing what we're working on in an easy to digest annual letter to share with friends and colleagues.

submitted 57 days ago • 1 comment

We have a new paper explaining all the ways you can use natural language processing to analyze text data in @natrevpsych.bsky.social We provide user friendly recommendations for using NLP to ensure rigour and reproducibility Here is a free link: www.nature.com/articles/s44...

submitted 55 days ago • 4 comments

Why do people distrust government agencies, even when they provide vital info (e.g., health, environment)? Can low-cost interventions restore trust & promote welfare-enhancing behaviors? Our study provides novel experimental evidence on these key questions (1/5) 👇 #EconSky

submitted 56 days ago • 1 comment

Large scale US-wide randomized field experiment by @hatethegamebook.com et al suggests flood risk information influences real estate consumer behaviour at each stage of the process, notably search, bidding and closing: https://buff.ly/3DAj58x via @mattkahn1966.bsky.social HT @RMetcalfe.bsky.social

submitted 58 days ago • 0 comments

This is a fantastic use of a behavioral map that shows the barriers to a key behavior like installing a heat pump, while also using comparison to hot tub and gas boiler installs to highlight just how much more complex heat pump installs are.

submitted 60 days ago • 0 comments

Research suggests a committed minority can overturn an established norm with a critical mass of ~25% of the population (> the 10% other studies have argued), showing good agreement between experimental and theoretical predictions for the tipping point: https://buff.ly/47zz7Jw

submitted 63 days ago • 0 comments

Nothing like waking up from a dream with an idea for an intervention. There are some nice papers that look at the effect of anti bullying interventions in schools where the authors intentionally focus on key nodes in the social graph. The idea being that if you have scarce resources that it makes

submitted 63 days ago • 1 comment

The paradox of effort—associated with frustration and avoidance, effort is also pursued for its perceived and retrospective value. @minzlicht.bsky.social et al suggest that,—despite its costs—it is critical in promoting persistence, growth and meaningful engagement: https://buff.ly/3Pe4mTj

submitted 64 days ago • 2 comments