In this purely theoretical work, I critique proceduralism in game study, which investigates the relationship between meaning-making in gameplay and game systems. I propose Gilles Deleuze's conception of identity and difference to construct a new image in which meaning is fundamentally unstable.

I explore the structural discovery potential of a particular agent-based system named Monte Carlo Physarium Machine - an algorithm inspired by slime mold. I apply it to language embedding data reduced to three dimensions. I also developed a visualization tool named Polyglot, which is available to use online.

We developed an AI agent that combined rule-based system and Monte Carlo Tree Search in a game environment similar to Bomberman. The highlight contribution of this project is the process of how we configure the algorithms to work in a resource-intensive environment.

I participated in a project that investigates confusing elements that can occur in C code. The confusing elements are termed atoms of confusions and their occurrances in software projects were studied using quantitative methods.