Abstract Floridi (Philos Technol 38(3):93, 2025a) has recently conjectured a formal trade-off between epistemic certainty and mapping scope in AI systems, suggesting a universal constraint that reappears in various guises from symbolic reasoning to statistical learning. We examine his proposal and identify several problems with the formalization, arguing that it obscures important distinctions between different kinds of AI systems and the tasks they address. More importantly, we contend that no formulation of the certainty-scope trade-off could be simultaneously precise enough to be tested and general enough to apply across different domains. The heterogeneity of AI itself ensures that any statement sufficiently abstract to encompass all cases must sacrifice the very structural detail required for a rigorous formalization. Rather than seeking an all-purpose law, we advocate a pluralistic, context-sensitive approach that treats the certainty-scope tension as a useful philosophical lens while resisting the impulse to over-formalize it.
Watson et al. (Tue,) studied this question.