CONCEPT
The Master Algorithm
Pedro Domingos’s hypothesis that a single universal learning algorithm exists—one from which all knowledge can be derived from data—and that the
five tribes of machine learning are five partial views of it.
The master algorithm is the hypothesis at the center of Pedro Domingos’s career: that there exists, or could exist, a single universal learning procedure from which all knowledge, past, present, and future, could be derived from data. The claim is not that such an algorithm has been found—Domingos is explicit that it has not—but that the apparent diversity of
machine learning’s five tribes is evidence that the field is circling a single underlying structure from five different angles, the way Maxwell’s equations emerged from what had appeared to be three separate phenomena. The master algorithm would be to machine learning what the standard model is to particle physics or the central dogma is to molecular biology: a unifying principle that explains apparent diversity as variations on one underlying mechanism. Domingos’s own candidates—Markov logic networks in the 2000s, tensor logic in 2025—are partial realizations of the vision: proofs that tribal fusion is achievable, not yet the fusion in full. But the hypothesis also forces