#AcademicRunPlaylist - 3/19/24

It was a pretty packed day for me, but I was still able to fit in some talks for my #AcademicRunPlaylist!

First was a great talk by Anita Williams Woolley on the foundations of group collective intelligence at the Carnegie Mellon University Software and Societal Systems Department https://www.youtube.com/watch?v=6OOOSHbhKt4

Next was an important talk by Joseph Scott Gladstone on historic business activity by native people in the western US at the Academy of Management. Gladstone reviews the anthropological and historical evidence to demonstrate the deep entrepreneurial and business roots in the native community leading up to and through relationships with colonial powers https://www.youtube.com/watch?v=OO6uJSVZaIY

Next was a wide-ranging talk by Pierre Andre Chiappori on human capital and heterogeneity within households at the Toulouse School of Economics https://www.youtube.com/watch?v=9tW_60x-0as

Next was an excellent talk by Tom Griffiths on using Bayesian models to understand large AI models at MIT Brain and Cognitive Sciences. There's a bit of hype at the beginning, but if you push through there's some interesting empirical work on why LLMs and other models fail (also for some reason this is a VR video) https://www.youtube.com/watch?v=yX_Az0VPVoY

Next was an intriguing talk by Krzysztof Gajos on the state of HCI knowledge in human-AI interaction at Stanford University. Gajos reviews the last 15 years of development in this space, pointing out how little researchers have interrogated interactions between people and algorithms https://www.youtube.com/watch?v=JybLJiJpKS0

Next was an engaging conversation with Rory McDonald on entrepreneurship, stakeholder theory, and more on the Stakeholder Podcast https://stakeholdermedia.libsyn.com/rory-mcdonald

Last was a fantastic talk by Simon Lucey on developing priors for neural networks at the Carnegie Mellon University Robotics Institute. Lucey provides great perspective on recent developments in AI and where these developments have been lacking, presenting compelling evidence that a neural prior approach can help further advance the field. Highly recommend https://www.youtube.com/watch?v=JOVPWLBIo5Q