Posted: 1/5/2019 Updated: 1/31/2019
Update Of 1/31/2019
Just watched part of this tutorial session presented at the NIPS (NeurIPS) 2018 conference titled Counterfactual Inference by Susan Athey (The Economics of Technology Professor at the Stanford Graduate School of Business). She actually talked about the intersection of economics and AI. Among other things she mentioned that Google used some game theoretic or other economic approach to develop AlphaGo.
Original Post
These are some personal, subjective observations. I have studied and graduated in economics.
Now that I have studied tons of research papers in machine learning and artificial intelligence, I can say that:
- Many of the statistical approaches used in AI for datasets or models have been used in econometrics for decades. Thus data science, machine learning, and AI have a very familiar ‘feel’ to me
- Economists have for decades worked with humongous datasets and massive models to predict human behavior and economic outcomes. Thus, I suspect that the recent, sometimes spectacular successes of AI are also founded on the previous research done in econometrics
- Researchers in AI do not make any cross references to any achievements of econometrics or applications of econometrics. Is this glaring omission a blind spot or some kind of ignorance?
Unfortunately, I do not have the time to explore this cross fertilization from econometrics to AI much further here. I will try to add to this post as time goes by
No comments:
Post a Comment