John Fox

Deontics Ltd., OpenClinical CIC

John Fox trained with AI founders Allen Newell and Herbert Simon, followed by a career researching the nature of human expertise, and applying this understanding to design of effective and versatile AI technologies. Focusing on medicine he has led work at MRC, CRUK, Oxford University and UCL/Royal Free Hospital, and major UK and European programmes. “CREDO” is a uniquely successful AI/knowledge engineering technology with many operational applications with 15+ trials reported in mainstream medical and technical journals. John has led several start-ups, including InferMed (acquired by Elsevier); Deontics Ltd, a world leader medical knowledge engineering, and OpenClinical CIC which offers a new way of developing and deploying AI services at scale. OpenClinical is a non-profit business whose mission is to empower healthcare professionals to develop AI services which are appropriate for clinical settings, directly addressing the safety, ethical and other issues which limit its practical adoption.

dont miss

Ensuring patient safety and ethical use of AI in practice

The current wave of excitement about AI focuses on machine learning but in the last wave (roughly 1980-2000) “expert systems” mostly used symbolic and logical models of (medical) expertise. Furthermore, learning methods also focus on creating “classifiers” while symbolic methods have a wide repertoire of capabilities, such as knowledge-based reasoning and decision-making, constraint based planning and design. A well-known reservation about symbolic methods, however, was that they are “brittle” in the face of real-world variability and uncertainty, a problem that probabilistic decision-making and learning are intended to solve. However, big data analytics and learning algorithms have their own issues, including undetectable biases and unaccountable “black boxes” that doctors and their patients cannot understand. This talk will outline how methods of data science can be integrated with those of knowledge engineering to yield versatile AI services, flexible care pathways and transparent and trustworthy decision-making.


  • Dr Stephen Weng: Speaking at the AI and Machine Learning Convention

    Dr Stephen Weng
    School of Medicine, University of Nottingham

    Machine-learning in prognostic research using routine health care data

  • Piotr Giedziun & Michał Krasoń: Speaking at the AI and Machine Learning Convention

    Piotr Giedziun & Michał Krasoń

    Deep learning in oncology diagnostics

  • Professor Claudia Pagliari: Speaking at the AI and Machine Learning Convention

    Professor Claudia Pagliari
    The University of Edinburgh

    Realistic AI - Understanding Potential in Context

  • Dr Matthew Fenech: Speaking at the AI and Machine Learning Convention

    Dr Matthew Fenech
    Future Advocacy

    Regulation of AI in healthcare: what should we expect?

  • Gael Kuhn: Speaking at the AI and Machine Learning Convention

    Gael Kuhn

    Advancing the development, utilization and clinical acceptance of AI