๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง
@m_sendhil
Follow People
On Danny Kahneman and 'angry science.' Samantha Power Arabella Pike Barack Obama John Tasioulas Richard H Thaler Piper Perabo Dilip Soman ๐จ๐ฆ Liam Delaney ๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง affective.brain Stephen E. Sachs Mia Farrow ๐ณ๏ธโ๐ ๐ป๐บ๐ธ๐
nytimes.com/2024/04/01/opiโฆ
Many of you will share tributes today to Danny Kahneman, highlighting how he informed your work or paved the way for your career, and so on.
Today I would like to honor DK for financing my relationship and eventual marriage to Carey Morewedge (1/n)
So excited and so very humbled to be stepping in to head AI Safety and Alignment at Google DeepMind. Lots of work ahead, both for present-day issues and for extreme risks in anticipation of capabilities advancing.
I'm hiring a #predoc to work with me at Harvard finance/macro/behavioral, and/or on Machine Learning (also with my great coauthors Ralph S.J. Koijen Motohiro Yogo Rob Richmond), starting in the summer 2024! Apply here: rb.gy/s2zwo3
Econ RA Listings PREDOC.org
Language modelsโ pretraining data includes information about historical events
When we use LLMs to make economic predictions, do they leak information about the future?
Keyon Vafa and I develop tests that find evidence of lookahead bias in LLMs, and identify ways forward (1/n)
Check out our sweeping 5-episode story on the growth and influence of behavioral economics
I was excited to produce this as a Opinion Science (a podcast)-Behavioral Grooves Podcast collaboration
The full series is here (opinionsciencepodcast.com/episode/they-tโฆ) + the Opinion Science or Behavioral Grooves podcast feed
Are algorithms overhyped? New research from Harris Policy's Jens Ludwig, Chicago Booth's ๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง, & Ashesh Rambachan on the effectiveness of algorithms for addressing public policy problems suggests the opposite:
ow.ly/HhFZ50QGI9o
๐๏ธThis Wednesday: the SRI Seminar Series welcomes Jon Kleinberg of Cornell Computer Science for a talk on user preferences and UX. Feb. 14, 12:30pm ET.
๐กTalk: โThe challenge of understanding what users want: Inconsistent preferences and engagement optimization.'
๐uoft.me/abR
Calculating the social return on algorithmic interventions, specifically their Marginal Value of Public Funds, across multiple domains โ regulation, criminal justice, medicine, and education, from Jens Ludwig, ๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง, and Ashesh Rambachan nber.org/papers/w32125
New working paper, 'The Unreasonable Effectiveness of Algorithms,' by Harris Policy's Jens Ludwig & Chicago Booth's ๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง calculates the social return on algorithmic interventions across multiple domains of interest to economists.
A new frontier in behavioral science: Why do we stop noticing things that are great? Why do we stop noticing things that are horrible? Dilip Soman ๐จ๐ฆ Paul Krugman Aimee Mann Claire Lehmann John Tasioulas ๐๐๐ง๐๐ก๐ข๐ฅ ๐๐ฎ๐ฅ๐ฅ๐๐ข๐ง๐๐ญ๐ก๐๐ง
amazon.com/Look-Again-Powโฆ
When January 1st falls on a Monday itโs a double fresh start. Read about the science of New Yearโs resolutions and the broader phenomenon they represent in this pair of research papers by Hengchen Dai, Jason Riis and yours truly. #NewBeginnings2024