Ben WaberLOTS of waiting for medical appointments today and one slow boot-walk mean lots of talks for my #AcademicRunPlaylist! (1/12)
First was an enjoyable discussion with Nina Lea Oishi and Caroline Grueskin on racial equity initiatives in the bond market and #OSHA's response to #COVID19. Both conversations extremely insightful, highly recommend https://podcasts.google.com/u/1/feed/aHR0cHM6Ly9hbmNob3IuZm0vcy8zODRlNWMyYy9wb2RjYXN0L3Jzcw/episode/OWNhNzlkZjItMTI5OS00N2YwLWI0YzctYjBmYTJlYzhhYzg5?sa=X&ved=0CAUQkfYCahcKEwjI3snyr-79AhUAAAAAHQAAAAAQLA (2/12) #ESG #DEI
Digging a Hole: The Legal Theory Podcast - Live Show!
For the last episode of the season, Digging a Hole hosted its first live show in front of a live student audience! David interviewed two current Yale Law School students – Nina Oishi and Caroline Grueskin – about recent papers they’ve written. We started by asking Nina about her recent paper “Rating Racial Equity: Examining BlackRock and Goldman’s New Racial Equity Initiatives in the Municipal Bond Market.” Yes, there are people other than David who write about the munibonds! Next, we talked to Caroline about her paper “At Least As Effective: OSHA, the State Plans, and Divergent Worker Protections from COVID-19.” We get deep in the weeds on OSHA and administrative law, and now you can too! Lastly, we brought an additional audience member in to do a “overrated” and “underrated” game with our interviewees. Come for the great student scholarship, stay for their hot takes. If you have any guest ideas for next season or ideas for a future live show, shoot us an emai…
Next was a fabulous podcast with @psmaldino and @add_hawk on the issues with value metrics and governance at the @sfiscience. There's deep conversation here of how KPIs can distort understanding, incentives, and who succeeds, and while I'd like to see more discussion of this in a non-scientific context (you need some coherent KPIs in companies) I still highly recommend it https://www.youtube.com/watch?v=tD6hh3ZAxWE (3/12)
Paul Smaldino & C
Next was an excellent discussion with @tomgoldstein on watermarking #LLMs on the #TWIML podcast. Goldstein presents an elegant method for watermarking LLM text to make it detectable, and hopefully this is widely deployed. Highly recommend https://www.youtube.com/watch?v=QC6wXsusHjk (4/12)
Watermarking LLMs to Fight Plagiarism with Tom Goldstein - 621
Next was a nice talk by Amit Sheth on combining symbolic #AI and #LLMs at the #USC Information Sciences Institute. As LLMs start to hit their limits, approaches like this will be necessary to continue improvements in AI https://www.youtube.com/watch?v=xyxQXka6dRY (5/12)
From NLP to NLU: Why we need varied, comprehensive, and stratified knowledge (Neuro-symbolic AI)
Next was a raucous talk by Raymond Mooney on answering why questions about narrative text at the #USC Information Sciences Institute. #LLMs still aren't great at this, but with some commonsense knowledge DBs and human evaluation described here significant performance is possible https://www.youtube.com/watch?v=mELMKmD2X1U (6/12) #AI #NLP
Answering Why Questions About Narrative Text
Next was an informative panel on the future of work at the #AoM with Jochen Menges, Matt Beane, Melissa Valentine, and Hila Lifshitz- Assaf. Matt didn't want to name that-which-shall-not-be-named, and I really enjoyed how the conversation spread beyond a lot of the hyped topics of the day https://www.youtube.com/watch?v=ZobfRnxhGec (7/12) #work #management
AOM Scholars On… Top 2023 Trends in the Future of Work
Next was a short talk by Claire McKay Bowen on providing different differentially private stats and analyses for tax #data. This was a wonderful breakdown of how to thoughtfully release meaningful, sensitive data to researchers. Highly recommend https://www.youtube.com/watch?v=ENvvF6jY8f4 (8/12) #statistics #DifferentialPrivacy #privacy
Claire Bowen: Fesibility Study of Differentially Private Summary Statistics and Regression Analyses
Next was a brief talk by Nathan Kurtz on how #data accelerates scientific discovery at the University of Washington eScience Institute. The examination of early #astronomy is incredibly relevant to #science today, and this is a great tour of how data transformed that field https://www.youtube.com/watch?v=G4m2iSl1EYM&t=2s (9/12)
Discovering AI@UW 2022 - Nathan Kutz - Data-driven accel. of scientific & engineering discovery
Next was an incredible talk by Ana Iltis on ethical issues in race-targeted genetic testing at #UNCC. #Race is often used to generate testing requirements for people to receive or donate kidneys, despite it being largely disconnected from genetics, & this important talk gets at the many ways that's problematic. Highly recommend https://www.youtube.com/watch?v=Zg5wLqh9cFU (10/12) #ethics #bioethics
"Race, Genes and Kidney Transplantation: Ethical Issues in Race-targeted Genetic Testing" -Ana Iltis
Next was an interesting talk by Debaditya Roy on predicting human behavior at #IITHyderabad. Roy nicely gets into the importance of assessing goals to accurately interpreting behavioral #data to predict behavior https://www.youtube.com/live/mtuqGcF6IGQ?feature=share&t=6484 (11/12) #MachineLearning
INCAPS-2nd Indo-Norway Workshop on Smart Sensing, Comm. and ML for Autonomous and CPS
Last was a detailed talk by Misha Khodak on #algorithms with predictions at #CMU. This gets back into CS algorithm theory, but focuses on more practical problems in learning and #privacy https://www.youtube.com/watch?v=aKL9pZB-nY0 (12/12)
