During the Autumn semester of my 3rd undergraduate year, I had a course that completely shaped how I think about the world and myself: Games and Strategic Behavior, taught by Alan Acosta. Shoutout to him.
Game theory was not just a subject I had to pass, but rather an acquired instrument that helped me conceptualize interactive dynamics, quantify preferences, valuations and beliefs, pinpoint prerequisites and underlying assumptions, model choices and incentives, ponder endogenous and exogenous factors, and find equilibrium scenarios.
Ironically, this economics-heavy subject came to my life just after changing majors from a BA in Economics to the BA in International Relations. And, I believe, it had a profound impact on my analysis of global affairs.
In fact, it prompted a question that would keep knocking at the back of my mind: if institutions are defined by human systems, would it be possible to reliably model their behavior? That would certainly require a comprehensive treatment of “noise” (assumptions from lack of information) and what may initially be seen as irrational actions (conceptualizing them rationally under other interpretations and frames). If this initiative worked, it would certainly mitigate the wild variability of thought that comprises IR studies due to having to put on theory-bound lenses to observe and understand reality. And those school-based perspectives often have an issue: they're not universal (even though some may claim to), and their effectiveness ends up depending on the characteristics of the evaluated actors or the type of conflict that's being encountered at the time.
That project would have been too big and idealistic for me to tackle from my position at the time. So I decided to start small, with something that could actually be shipped:
Parting from the implementation of a new global policy, can we leverage observed results to understand the actions, preferences, belief-weighted valuations, and power dynamics of the participating agents?
This is one of the questions that shaped my undergraduate thesis and, as you can see, it's strikingly similar to the questions that mark 3 of my 4 current research tracks.
- In the Psychometrics track, we understand the “policy” as an episode of an administration regime, and the “participating agents” as the assigned roles in the experiment.
- In the Societal Safety track, we keep essentially the same question, but understand the nature of implementators in a broader manner than traditional institutions.
- In the Interactive Societies track, we keep that same question too, but care more about the relations between agents rather than the administered policy itself.
Then there's the legal track, which doesn't really have a connection, as it arose from my work experience and was then consolidated from my involvement at FAIR. If we try to find the connection anyway, we could argue that in that case we'd replace “new policy” with “reasoning constraints” and care more about detecting variance rather than explaining it, which marks a key difference. There, the goal is not to understand how each parameter influences the final output, but instead to measure the capabilities of a system to execute an objective.
Looking back, it's clear that while these matters may not be perfectly aligned, they can certainly be inserted in a coherent grammar by learning to speak in different vocabularies. That's what I love about interdisciplinary adventures. That's my specialty.