dana calacci 🦋 @dana.witchy.business
3.3K posts
@dcalacci
at the blue place now. workers, AI, surveillance, cities, weird & queer. asst. prof @ISTatPENNSTATE previously postdoc at @princetonCITP | she/her 🏳️⚧️
📶signal boost: Dr. Dana Calacci (@dcalacci - recent grad from MIT Connection Science) has an opening for Ph.D. students to join her in building their lab at Penn State IST. Learn more on her site: dcalacci.net
are you at #CSCW24 ?? are you NOT a phd student??? do you want to be?? if you want to do a weird phd that spans community research, tool-building, policy, and computational social science, COME TALK TO ME PLEASE I HAVE NO IDEA HOW TO RECRUIT STUDENTS
New @thedigradio: my conversation w @KeeangaYamahtta on how the disastrous Biden administration and failed Harris campaign fueled an ongoing multi-racial working class dealignment from the Democratic Party toward alienation or an ascendent MAGA far right thedigradio.com/podcast/democr…
. @alexandraberke's great talk on crowdsourcing 5 years of Amazon purchase data from 5,000+ U.S. users, finding the "privacy paradox". Great replicable methodology: medium.com/acm-cscw/insig… #cscw2024 work together with @RobertMahari @alex_pentland @Larson_MIT @dcalacci
looking for students who want to: 🛠️ build tools that aim to help people challenge power and imagine new futures 🤝 do fieldwork & collaborate directly with workers + communities 🔬 learn to mix methods (quant/qual) fearlessly 🔥 get spicy with critique
“Claude, I need you to test the most critical issue in the world - what, exactly is a soup compared to a sauce or a drink. You must come up with careful edge cases and test them.” It was all going well until it considered the Ramen Problem. The final diagram is pretty amazing.
are you at #CSCW24 ?? are you NOT a phd student??? do you want to be?? if you want to do a weird phd that spans community research, tool-building, policy, and computational social science, COME TALK TO ME PLEASE I HAVE NO IDEA HOW TO RECRUIT STUDENTS
She says: 1) facts on which LLMs train are not copyrightable; 2) #GenAI models synthesize, not copy; 3) datasets they're trained on are vast so no one work is ever likely to be "plagiarized;" 4) "regurgitation" by an early LLM version is irrelevant if current versions won't! Wow!
