Amit Gandhi
@AmitEcon
Followers
745
Following
841
Media
0
Statuses
198
Vice President and Technical Fellow at Airbnb (former Chief Economist at Microsoft Azure)
Philadelphia, PA
Joined June 2013
I am super hyped to finally share the first release of plumber2 with all of you. This has been the center of my attention for a big part of 2025 and I hope you'll find it a worthy update to the venerable plumber package. #rstats
https://t.co/mGMk2SwuWz
0
10
40
GNRProdEst.jl Update! Main new feature: Bootstrapped Standard Errors! I just released the latest update of my @JuliaLanguage implementation of the production function estimation method from @AmitEcon, @snavarrol, Rivers (2020). Check it out here
github.com
Implementation of the Gandhi, Navarro, Rievers (2020) production function estimator. - MarkusTrunschke/GNRProdEst.jl
1
5
11
One person I feel who is insufficiently appreciated is Hadley Wickham. The results of his work are everywhere, every journal and conference I know. He changed the history of science.
11
102
711
GNR took 12+ years to complete from idea to publication. During that time Stata remained the required language to communicate with the profession. Amazing to see how far the field has come with open source and modern software standards. Thanks @MTrunschketo - can't wait to use!
Btw, I recently finished my @JuliaLanguage implementation of the @AmitEcon , @snavarrol, Rievers (2020) production function estimation method. It's a now a registered Julia package! Go check it out here https://t.co/LkklbZjjVP Feel free to contribute to dev and improve it.
2
5
24
Couldn't agree more @LeistenEcon :) In all seriousness - zero shares are rampant inside of business environments and the corrections have first order effects on the insights derived. Fine tuning the priors in an Empirical Bayes kind of way sounds fascinating - thx for sharing.
The Gandhi et al. (2023) approach to zero empirical shares is so intuitive & I don't know why it isn't more widely adopted. Like, of course s_ij = 0 because true choice probs are positive but small and we only observe finite demand draws! https://t.co/BX3SGxnwHG
0
0
19
"Perfection is impossible. In the 1,526 singles matches I played in my career, I won almost 80% of those matches. But what percentage of points did I win? 54% In other words, even top ranked tennis players win barely more than half the points they play. When you lose ever
425
12K
86K
The world's first 60+ FPS e-ink display by @daylightco on Episode 45 of S³ See how it works, the 6-year development journey, and Daylight's vision for the future of personal computing.
80
249
3K
Excited to discuss the state of data and tech with my former colleague Francesco Decarolis - thanks to the organizers for the invitation.
... Francesco Decarolis & @AmitEcon, Pietro Garibaldi & Antonio Spilimbergo, Stefano Corgnati & Marco Ottaviani, Michael Jordan & James Manyika & Michael Spence, David Card & @Danielle__Li & @wellingmax, and many more... 7/
0
0
3
One of my favorite tools for visualization of big data is tippecanoe, which converts millions of spatial features (like building footprints) into scale-sensitive vector tiles https://t.co/kZM1DKt2cg You can use tippecanoe in #rstats via its wrapper in the mapboxapi package:
3
21
156
Thrilled to see the innovation unfold at the #Wharton Hack-AI-thon! Best of luck to all participants! #AI #Innovation #AnalyticsAtWharton
Excited to kick off our inaugural #Wharton Hack-AI-thon with a special thanks to @AIatWharton faculty co-director @StefanoPuntoni for starting the day! We can't wait to see the innovative apps & prompts from our talented #students. Best of luck to everyone! #AI #Innovation
0
1
9
We're thrilled to welcome @AmitEcon and Cornelia Walther to our team as Visiting Scholars! Their expertise in corporate partnerships and leveraging technology for pro-social change will help take our analytics mission to greater heights. Learn more here: https://t.co/hVklzz5tZS
0
2
2
Excited to share our work w/ @hamsabastani and @obastani as my debut post at X!
Learn how @Wharton PhD student, @AngelTHChung, leveraged machine learning to help the government of Sierra Leone distribute limited medical aid to the communities who need it most, as we continue to highlight Wharton Women in Analytics. https://t.co/FetlwQ4WC6
0
2
9
Teams of high school students from across the globe are crunching numbers and competing in the @Wharton High School Data Science Competition with @WhartonSABI and @whartonyouth. Meet some of the teams competing to make the best data predictions about our fictional soccer league!
0
7
8
"Causal questions are in 1st order in tech firms." "Lots of good executives have a causal theory of reality in their heads. What they often want to see is... dashboards, metrics... these highly correlated data where the causal filtering is happening in a board room, in a
0
1
4
"Don't care" might be a little strong - but the need to more deeply understand the ATE and the "why" behind its movement is paramount. That leads naturally to questions of heterogeneity and more fundamentally the structural mechanism (great opportunities generally for economics).
0
1
25
This paper by top AI researcher @sleepinyourhat on what is surprising about the new AI Large Language Models is an absolute must-read. A lot of what is happening was not on anyone's radar. What will happen next is hard to predict. https://t.co/IlXHfVArOL H/t @emollick
1
5
50
As ChatGPT and other LLMs become better and better at *answering questions,* the important complementary human skill becomes: Asking questions. As I tell my students... Asking really good questions is an art form that is becoming even more important in the Age of AI.
9
54
234
Hotter take: ML would have advanced faster if another front-end language had been available and widely adopted instead of Python. One that is interactive yet fast & compilable, multithreaded (no GIL), isn't bloated, doesn't care about white spaces,... E.g. Julia or some Lisp.
Hot take: Machine Learning would not have been nearly as advanced now were it not for Python. Python’s two main virtues in the context of ML: 1. Lowering barriers to entry. 2. As a scripting language, it encourages and enables experimental workflow.
296
370
3K
Has Microsoft put Google on a defensive foot wrt to their search engine dominance? What do you think?
1
1
8