Shared worlds, shared minds
This article explores the training of embodied AI systems and robots through engagement with physical and social environments, drawing parallels with living organisms. We examine key challenges in developing embodied AI systems that learn through prediction, interaction and modification of their environments and discuss the societal implications.
Authors
- Thomas Parr (ORCID: https://orcid.org/0000-0001-5108-5743)
- Giovanni Pezzulo (ORCID: https://orcid.org/0000-0001-6813-8282)
- Karl Friston (ORCID: https://orcid.org/0000-0001-7984-8909)
Institutions
- Queen Mary University of London (GB)
- Nuffield Orthopaedic Centre (GB)
- University of Oxford (GB)
- Oxford BioMedica (United Kingdom) (GB)
- National Hospital for Neurology and Neurosurgery (GB)
- National Research Council (IT)
- Institute of Cognitive Sciences and Technologies (IT)
- University College London (GB)
Publication Details
- Journal
- EMBO Reports
- Published
- 2025-08-20
- DOI
- https://doi.org/10.1038/s44319-025-00549-8
- Citations
- 1
- Primary Topic
- Embodied and Extended Cognition
- Type
- article
- Field-Weighted Citation Impact
- 1.43
Funders
- Wellcome Trust
- European Research Council