Writings
(Last updated: November 2025)
I wrote this essay as part of a seminar course I ran in the Fall of 2025 on the economics of AI.
Abstract: The purpose of this brief note is to identify unique microeconomic properties of humans. Traditional economic analysis focuses on productive activities (or tasks), leaving both game-theoretic aspects and explicitly socio-psychological aspects outside the frame. We aim to highlight microeconomic properties unique to humans that remain relevant even if any task (including long-term planning and ethical judgement calls) which humans can perform can also be performed by machines. By necessity, all these are based on what humans cannot do (or cannot have done to them), or on the way humans uniquely care about what other humans think or feel. We break features into these two categories.
Some of these differences can be remediated using legal tools (an example of such a tool is the legal fiction of a limited liability corporation), should society choose to do so, but such a choice is not in any way technologically predetermined (unlike achieving “AGI”, which is ultimately mainly a technological problem).
Full essay: [google doc]
(Last updated: December 2020)
Abstract: There is a growing power imbalance between companies who have access to powerful algorithms and processing capacity and users who don’t. To restore the balance, we need to put equally powerful algorithms in the hands of individuals. There are two concrete steps we recommend:
(1) there should be a new regulatory framework which creates an unbreakable commitment for an advocate (a digital service) to work exclusively in the interest of its client;
(2) existing and new regulations around digital rights of individuals (such as GDPR) should make it a priority to make it easy for users to take advantage of these rights using software.
Full essay: [pdf]