Hannah Li

I am teaching a PhD seminar this fall which attempts to put a decision-theoretic lens (or lenses) on research design.

So e.g., statistical decision theory, eliciting beliefs & preferences, and rational benchmarks.

Here's a draft syllabus docs.google.com/document/d/1...

What's missing?

Joint work with Liang (Charles) Lyu, James Siderius, @dacemoglumit.bsky.social, Dan Huttenlocher, and Asu Ozdaglar

We see a similar effect in edits, although not statistically significant. Very important to keep monitoring in the future, as platforms like Wiki serve as important info repositories (and training data for future LLMs)

This is in contrast to prior work that sees no significant change to Wikipedia articles *as a whole*, after the launch of ChatGPT. This masks selective changes in consumption patterns.

Wikipedia Contributions in the Wake of ChatGPT How has Wikipedia activity changed for articles with content similar to ChatGPT following its introduction? We estimate the impact using differences-in-differences models, with dissimilar Wikipedia ar...

Just out in WWW last week! 📜Our work on substitution patterns between Wikipedia and ChatGPT. We find *heterogeneous* impacts, where Wiki articles that are similar to ChatGPT outputs see a greater drop in views than dissimilar articles:

arxiv.org/abs/2503.00757

I am influenced. Buying a copy too!

Curious to see what analyses you do! I've been tracking my habits and moods on Daylio for the last 2 years for no reason but to one day do some data analysis on myself... maybe that day will come soon

Come work with people like @dholtz.bsky.social, Nicola Rosaia, and me :)

📣 Please repost 📣 We are hiring a staff Research Associate (pre-doc) at Columbia DRO!

This is a great opportunity if you want to work on online platforms, urban economics, experiment design, or impacts of AI.

Deadline to apply is May 15. For more info: business.columbia.edu/faculty/divi...

Terrifying how much control the Trump administration has on our education system...

*Please repost* @sjgreenwood.bsky.social and I just launched a new personalized feed (*please pin*) that we hope will become a "must use" for #academicsky. The feed shows posts about papers filtered by *your* follower network. It's become my default Bluesky experience bsky.app/profile/pape...

I definitely believe these trends are happening, but I can't figure out how they would reliably measure this? Their method relies on classifying which texts are AI generated, and so the classifier could just be flagging texts with low diversity

Wow now THIS is some cool data viz. Insane looking at Manhattan and seeing the Billionaire's Row ultra high rises casting 20 block long shadows across Central Park...

Yes to evaluating the *outcomes* of these systems rather than as a standalone algorithm! This is something that's been bothering me for a while about ML assisted decisions

"Life can get very exciting upon realising that a problem has more {axes/limits/...} than was initially apparent." <-- Maybe this says something about how exciting my life is, but I related to this...

Real footage of a synthetic control model

Hmm what are you suggesting to divide by?

Cursor has been a lifesaver for me!!

neat - wonder if there's some way to disentangle how these starter packs have shaped the resulting follower-following graph, compared to if we only had organic search

This malaria vaccine is delivered by a mosquito bite Bites from insects infected with modified malaria parasites boosted immunity and stopped people from contracting the disease.

Coolest thing I've seen in a minute: hijacking mosquitoes to spread the malaria *vaccine* when they bite people

www.nature.com/articles/d41...

Would love to be added to this! :)

Excited for the dialogue and the research community that this site is going to create!

ICYMI a starter pack to fill your feed with awesome women in AI go.bsky.app/LaGDpqg

Can you add me as well? Thanks!