Nicholas Emery-Xu
@nicholaskemery
Economics Ph.D candidate @UCLA | economics of computing and AI | industrial organization, innovation, political economy | PEPFAR fanatic
One of the most terrifying facts is that these two, uh, stable, level-headed men control ~90% of the world’s nuclear weapons
I had a great time at @nber.org Digital Economics and AI tutorial and meeting. Met a lot of people working on very cool projects!
In Pokémon Gen V, Unova is so strongly based on New York that it even has its NIMBYs🥲
Enjoying the MIT FutureTech conference on AI automation in science and labor @mitsloan.bsky.social @csail.mit.edu
(However, according to data from cloud security firm Wiz, most firms tend to use a primary provider for most of their data) www.wiz.io/lp/state-of-...
A counterpoint to the slow diffusion of technology narrative is the extremely fast growth of infrastructure as a service (IaaS). In my data on large French firms, I find usage increases from 1.2% in 2017 to 22.3% in 2021! And a large minority of firms adopt multiple providers.
The labor share of income in US R&D has risen by 20pp in the last 30 years!
In contrast to prior results in Armstrong et al. (2016), when states are uncertain about how capabilities translate into success in the race (low m), they should make their capabilities publicly known to reduce risk. (5/11)
📣New paper with Andy Park and Robert Trager. We build a formal model to study the role of information in a risky technology race. We find that public knowledge of capabilities is safer than private when the contest is noisy and that risk-sharing increases risk. (1/11)