Interests

I am an economist by training and I spent a decent amount of time as an academic economist. To an outsider, my research might seem all over the place, from public economics, behavioral economics, to mostly computer science + game theory.

To be honest, during my PhD I was maendering through different topics and methods and that is how my publications look like. I was also not very much convinced of the usefulness and relevance of my home academic turf - economics. It took quite a while until I figured out what I actually care about: It is the art and science of effective organization.

By that I do not meant what is sometimes claimed by academic departments as their turf, but I mean the broad interest in understanding how we (the humans, and well, by now also the AIs) can organize themselves better. Naturally, this includes technology.

In the overall evolution of societies and compared to technology, our progress there still seems underwhelming (and having had to apply for public research funding does not really positively enforce one’s beliefs).

Of course, I am not the first one interested in effective organization, and there was a whole movement who was interested in understanding these principles (there is a reason we called our research institute the Institute for Categorical Cybernetics). And yes, most of this has vanished and plays basically no academic role anymore.

So why stubbornly stick to this? For the last 10+ years, I have been worked on the intersection of economics and computer science. More precisely I was involved in the development of a new mathematical language for game theory (game theory is a language for reasoning about agents). The language we developed is based on category theory, an abstract form of mathematics orginally used in pure math, later in theoretical computer science and more recently also an applied modelling language in different fields.

The language we developed has two features: It is compositional and it can be used as a blueprint for an actual programming language. Which we implemented and have been using in 20squares as a business tool.

Working on this project meant crossing scientific disciplines and producing artifacts (code and tools) very different from what academics produce (papers). It also meant venturing into business (at 20squares consulting we use the tool we developed for our business) and adjacent social-technological worlds: I have worked on applications in the blockchain space, large world game design, consequences of algorithmic pricing, LLM interfaces for economic design, and AI agents frameworks.

All of this has reinforced my interests in the science and engineering of effective organization. And as to today decisions are mediated through software, supported by software, and increasingly made by software, for me this means an interest in understand and designing the “socio-technological” stack. And as building a stack means building frameworks and tools, I have spent most of my time doing that.

In practice, my focus is still often more narrowly on the on economic decision-making. In particular, how (parts of) economic modelling can be transformed from a more artistic and individualistic activity into an industrial process which scales.