Probabilistic Effectivity in the Subspace Theorem

Roth’s Theorem, together with its far-reaching generalization, Schmidt’s Subspace Theorem, are cornerstone results in Diophantine approximation, providing quantitative bounds on how closely algebraic quantities can be approximated by rational ones. These theorems have diverse applications — for instance, in the analysis of linear recurrence sequences, in the resolution of Diophantine equations, and in the study of the complexity of the decimal expansions of algebraic numbers. A central limitation of these results lies in their ineffectivity: the quantitative estimates they provide involve constants which are not effectively computable. In this talk, we present recent joint work with Faustin Adiceam in which these classical theorems are reformulated within a probabilistic framework. This perspective yields quantitative and effective statements that hold, in a suitably defined sense, almost always.