A virtual, data-driven replica of a physical object, system, or person that mirrors its real-world counterpart in structure, behavior, and state through continuous streams of sensor data, simulations, or recorded activity. Used in engineering, manufacturing, healthcare, and urban planning, a digital twin allows operators to test scenarios, predict failures, and optimize performance without directly intervening in the physical system itself. By modeling reality in software, it functions as a sandbox for experimentation, enabling safer decisions and more efficient design.
In contemporary AI and consumer contexts, the term has expanded beyond industrial machinery to describe increasingly personal or behavioral models, such as algorithmic profiles, simulated avatars, or lifelike generative agents that approximate an individual’s preferences, speech, or decision patterns. This broader usage frames the digital twin not just as a technical model but as a mirror of identity, raising questions about authorship, autonomy, and ownership as one’s “data self” becomes capable of acting, predicting, or even creating on one’s behalf.