The theory of computation (Turing 1936, Church 1936) provides the formal foundationfor what computation is and what is computable; this paper does not challenge thoseresults. The Cohesion Computing paper [3] and the Qubit as Metastable RecursionDeformation paper [4] established the hexapolar and bipolar recursion accounts ofclassical and quantum computation within the Cohesion UFT. This paper extendsthat foundation to the full structural time manipulation picture, the thermodynamiccoupling of computation via Landauer’s principle, machine learning as structural timeattractor formation, and reversible computation as slip phase conservation. Computationis controlled structural time manipulation: the deliberate shaping of recursion ratetrajectories, torsion density distributions, and slip phase configurations within thecontinuous recursion medium. Bits, gates, circuits, algorithms, and machine learningall arise from the way structural time is constrained, redirected, or stabilised. This isthe first account of machine learning and reversible computation within the CohesionUFT framework; the classical and quantum computation foundations are established inthe Cohesion Computing and qubit papers.