---
id: posts/sfi-building-better-digital-twins
about:
  label: Building better digital twins
  description: >-
    The Santa Fe Institute's report on its September 2026 working group: why the
    organisers think the moment has come, what a twin of a cancer patient would
    need, and why trust and validation will decide whether twins are used.
  tags:
    - complex systems
    - medicine
    - trust
    - verification and validation
    - Santa Fe Institute
article:
  url: 'https://www.santafe.edu/news-center/news/building-better-digital-twins'
  published: '2026-10-22'
  publisher: Santa Fe Institute
domains:
  - people
  - places
relations:
  - predicate: about
    target: events/sfi-digital-twins-for-complex-systems-2026
provenance:
  origin: agent
  sources:
    - title: 'Building better digital twins (SFI news, 22 October 2026)'
      url: 'https://www.santafe.edu/news-center/news/building-better-digital-twins'
  created: '2026-10-23'
  reviewed: '2026-10-23'
  reviewed_by: 'Claude, from the source listed'
  status: draft
---
The Santa Fe Institute's story on its [working group on digital twins](/events/sfi-digital-twins-for-complex-systems-2026/) lists the things a twin might stand for: a bridge, a spacecraft, a forest fire, a tumour, a city. What makes each a twin, in the story's terms, is that it keeps updating from the real thing, so interventions can be tried on the model first.

Karen Willcox, one of the organisers, says interest has surged and the timing is right, pointing to advances in AI and to interest from the National Institutes of Health, the National Science Foundation and the Department of Defense. The open problems the story lists are the cost of the computing, earning public trust, and deciding how much detail a model needs to be useful. Willcox puts trust first: whether twins can be made trustworthy, explainable and checkable will decide whether they are adopted.

Most of the story goes to medicine. Caroline Chung, a radiation oncologist at the MD Anderson Cancer Center, argues that a twin of a patient has to run from the genome upward, through metabolism and the immune system to how the person lives. A clinical trial usually compares one treatment schedule with standard care; a twin could ask which patients do better on more or less of a drug, and how often. The obstacles she names are data spread across scales, research groups working in silos, data that is missing or inconsistent, and keeping a model current as the science moves.

Her list fits a model of a place nearly as well. A watershed runs from a soil core to a river basin, and its data sits with separate agencies in separate formats. How much detail is enough is a question a small team meets early when it models its own valley.