AI vendors push prescriptive digital twins; researchers warn
Industry writing says digital twins for urban planning are shifting from description toward predictive and prescriptive functions, offering not just visualization but automated solutions. At the same time, UCLA researchers caution that AI systems lack bodily experience and awareness—an embodied dimension of human judgement that is hard to capture in models (futureinsights.com, smdp.com).
A digital twin is a live digital copy of a city, and vendors are now selling versions that do more than show streets and pipes: they recommend what officials should do next. (futureinsights.com) Future Insights wrote that by 2026 urban digital twins are moving beyond static three-dimensional models into real-time systems fed by Internet of Things sensors, satellite imagery and high-speed networks. The same article said planners can simulate floods, stadium traffic and other scenarios before construction starts. (futureinsights.com) That sales pitch is colliding with a warning from researchers at the University of California, Los Angeles. In a UCLA Health release published April 3, 2026, postdoctoral fellow Akila Kadambi and colleagues said current artificial intelligence systems still lack “internal embodiment,” or an internal sense of bodily state such as fatigue, uncertainty or need. (uclahealth.org) The researchers said humans do not make judgments from language alone. Their example was passing a saltshaker: a person tracks hand position, object feel and social context in one motion, while a model can describe the act without having lived it. (uclahealth.org, smdp.com) That gap matters in city systems because digital twins are being framed as decision tools, not just dashboards. Future Insights said the newer pitch is prediction and optimization, while a 2025 United Nations Development Programme brief described digital twins as tools for prioritizing resources under pressure from climate change, urbanization and strained infrastructure. (futureinsights.com, undp.org) European guidance for local policymakers published in 2025 also places urban digital twins inside evidence-based policymaking, procurement and community engagement. That means the systems are being discussed as part of how governments buy technology and justify public decisions, not only how they visualize data. (vcity.tech) UCLA’s paper does not say cities should stop using artificial intelligence. It says systems that sound fluent can still miss the bodily constraints humans use as a built-in safety check, especially when they are deployed in consequential settings. (uclahealth.org) The split is becoming clearer: vendors are marketing city twins as engines that can optimize traffic, energy and land use, while researchers are warning that optimization is not the same thing as judgment. For city officials, the question is no longer whether a digital twin can model a street, but how much authority to give a model that has never had to cross one. (futureinsights.com, uclahealth.org)
