Aaron Schecter (@aaronschecter) 's Twitter Profile
Aaron Schecter

@aaronschecter

Assistant Prof of MIS @ UGA. Interest in networks, AI, and stats. Proud nerd.

ID: 1905068233

linkhttp://aaronschecter.com calendar_today25-09-2013 18:27:25

572 Tweet

359 Followers

513 Following

President Biden Archived (@potus46archive) 's Twitter Profile Photo

Glory glory, @UniversityofGA. Congrats to the Dawgs on a hard-fought National Championship – and to TCU Football for beating the odds all season. Georgia, no doubt you made your community proud tonight.

Eszter Hargittai 😷💉🌻 (@eszter) 's Twitter Profile Photo

⭐ my department is hiring! ⭐ Tenure-track assistant professor in computational social science with a focus on communication University of Zurich is a wonderful place to work! jobs.uzh.ch/offene-stellen…

Erik Brynjolfsson (@erikbryn) 's Twitter Profile Photo

We live in interesting times. “These findings suggest that a “theory of mind”-like ability (thus far considered to be uniquely human) may have spontaneously emerged as a byproduct of language models' improving language skills.” arxiv.org/abs/2302.02083

Kareem Carr, Statistics Person (@kareem_carr) 's Twitter Profile Photo

I've spent hours experimenting with using multiple large language models (LLMs) such as Bard AI, GPT-4, GPT-3.5 and Bing AI for academic research. They're extremely powerful but with some very painful pitfalls. Here's my guide to using LLMs as your personal research assistant:

I've spent hours experimenting with using multiple large language models (LLMs) such as Bard AI, GPT-4, GPT-3.5 and Bing AI for academic research.

They're extremely powerful but with some very painful pitfalls.

Here's my guide to using LLMs as your personal research assistant:
Aaron Schecter (@aaronschecter) 's Twitter Profile Photo

Finally attending my first CHI conference! Loving the debates about what is new and what is not with modern AI tools #CHI2023 #sigchi

Ethan Mollick (@emollick) 's Twitter Profile Photo

Since I wasn't that clear, what made this cool was that I simply gave the AI a large (100MB+) dataset that had been used in a prior analysis, and told it to figure out what was interesting. It replicated the results of teh prior analysis in a pretty sophisticated way.