Digital Twin: Will a Perfect Virtual Copy of the World Help Us Predict the Unpredictable? 5 Limits That Even the Best Simulation Cannot Cross

This article examines digital twin technology — continuously updated virtual replicas of physical systems — from its conceptual origins in NASA's 1970 Apollo 13 rescue through its formalisation by Michael Grieves (2002) and NASA engineer John Vickers (2010), to its current applications at city scale (Virtual Singapore, the Singapore Land Authority's national 3D digital twin, operational since 2018) and organ scale (Dassault Systemes' Living Heart Project, a multiphysics cardiac digital twin used clinically since 2014, including in the documented paediatric case of patient Annika Seed at Boston Children's Hospital). Having established the evidence base, the article identifies five structural limits that prevent any digital twin from predicting genuinely unpredictable events: the impossibility of computationally replicating consciousness rather than merely its neural correlates; emergence, as formalised by Nobel laureate Philip Anderson in his 1972 Science paper 'More Is Different,' which demonstrates that higher-level system properties are not derivable from lower-level component rules regardless of computational power; the Black Swan problem articulated by Nassim Nicholas Taleb (2007), in which genuinely novel, high-impact events fall outside any dataset a model could have been trained on; the Vedantic concept of Maya, which holds that representation and reality remain categorically distinct regardless of representational fidelity; and the ethical limits of simulating human beings, including consent, privacy, and the bias risks documented in current human digital twin research. The conclusion identifies what digital twins are reliably good for — and where their reliability necessarily ends.