Kingson Man
Neuroscientist and AI builder
I study how feeling shapes intelligence.
At Feeling Machines, I build artificial agents with needs and vulnerabilities to explore how empathy and care might arise. My earlier work used brain imaging to study how sight, sound, and touch come together into a coherent world.
Recent:
The Science of Consciousness, San Diego, October 2026
ASSC poster & videos, Santiago, July 2026
Selected papers
Homeostasis and soft robotics in the design of feeling machines
Kingson Man & Antonio Damasio. Nature Machine Intelligence, 2019.
-
Preventing antisocial robots: A pathway to artificial empathy
Leonardo Christov-Moore et al. Science Robotics, 2023.
-
Homeostatic Coupling for Prosocial Behavior
Naoto Yoshida & Kingson Man. ALife, 2025; NeurIPS IMOL Workshop, 2024.
More publications
-
Need is All You Need: Homeostatic Neural Networks Adapt to Concept Shift
Kingson Man, Antonio Damasio & Hartmut Neven. NeurIPS NeuroAI Workshop, 2024.
Testing whether vulnerability and self-regulation help neural networks adapt when learned associations change. First released as a preprint in 2022.
-
Convergent and invariant object representations for sight, sound, and touch
Kingson Man, Antonio Damasio, Kaspar Meyer & Jonas T. Kaplan. Human Brain Mapping, 2015.
Mapping where information about objects converges across senses, and where representations generalize between them.
Homeostasis and feeling in artificial agents
- 2024Empathic coupling of homeostatic states for intrinsic prosocialityYoshida N, Man K. NeurIPS 2024 Workshop on Intrinsically Motivated Open-ended Learning.OpenReview
- 2021AI reflections in 2020Jobin A, Man K, Damasio A, et al. Nature Machine Intelligence 3.
- 2020Truth or consequences: Homeostatic self-regulation in artificial neural networksMan K, Damasio A. Proceedings of the Artificial Life Conference 2020, 146-147.video
- 2019Homeostatically motivated intelligence for feeling machinesMan K, Damasio A. AAAI Spring Symposium on Towards Conscious AI Systems. CEUR Workshop Proceedings 2287.
Neuroscience
- 2020Seeing objects improves our hearing of the sounds they makeMan K, Melo G, Damasio A, Kaplan J. Neuroscience of Consciousness 2020(1), niaa014.free text
- 2017Decoding the neural representation of story meanings across languagesDehghani M, Boghrati R, Man K, Hoover J, Gimbel SI, Vaswani A, Zevin JD, Immordino-Yang MH, Gordon AS, Damasio A, Kaplan JT. Human Brain Mapping 38(12), 6096-6106.free text
- 2015Multivariate cross-classification: Applying machine learning techniques to characterize abstraction in neural representationsKaplan JT, Man K, Greening SG. Frontiers in Human Neuroscience 9, 151.
- 2014Mapping the neural architecture of conceptsMan K. PhD dissertation, University of Southern California.
- 2013Neural convergence and divergence in the mammalian cerebral cortex: From experimental neuroanatomy to functional neuroimagingMan K, Kaplan J, Damasio H, Damasio A. Journal of Comparative Neurology 521(18), 4097-4111.free text
- 2012Sight and sound converge to form modality-invariant representations in temporoparietal cortexMan K, Kaplan JT, Damasio A, Meyer K. Journal of Neuroscience 32(47), 16629-16636.free text
- 2008Glutamatergic contributions to nicotinic acetylcholine receptor agonist-evoked cholinergic transients in the prefrontal cortexParikh V, Man K, Decker MW, Sarter M. Journal of Neuroscience 28(14), 3769-3780.
Other
- 2020Explainable non-cooperative spacecraft pose estimation using convolutional neural networksAlimo R, Jeong D, Man K. AIAA SciTech 2020 Forum. Work from the NASA Jet Propulsion Laboratory.
Research notes
Feeling, vulnerability, and artificial agents
My current work, Skin in the Game, investigates how embodied vulnerability changes behavior. In simulation, agents’ energy and water reserves affect their ability to act. I study how they respond to threats and anticipate the bodily consequences of their choices. Learning useful models of those consequences remains an open problem.
Homeostasis, learning, and cooperation
I study how sharing internal states can support helping behavior between simulated agents, and how neural networks that regulate their own learning rates adapt when learned associations change.
How the senses become a world
My doctoral and postdoctoral work used brain imaging and machine learning to study how the brain represents objects across senses. I also investigated how seeing an object changes our ability to hear it, and how the brain represents the meaning of stories across languages.
Elsewhere
- A World Appears. Featured in Michael Pollan’s book on consciousness (2026).
- Wider than the Sky. An appearance in Valerio Jalongo’s documentary on AI (2025). Film & trailer.
More interviews & coverage
-
RTVE, Spanish television
¿Puede una inteligencia artificial aprender empatía?
On bodies, vulnerability, and machine feeling, for El cazador de cerebros.
-
Science News, Tom Siegfried
A will to survive might take AI to the next level
On the proposal to build artificial intelligence around homeostasis and self-preservation.
- Michael Pollan: The Mystery of ConsciousnessOn Being, 2026, A brief mention of my work in Pollan’s interview with Krista Tippett
- Can we prevent AI from acting like a sociopath?USC Dornsife, 2026, Coverage of research on artificial vulnerability and empathy
- Robots Need to Know They Can Die at Any Minute, Just Like the Rest of UsPopular Mechanics, Caroline Delbert, 2019
- Found in translation: USC scientists map brain responses to stories in three different languagesUSC Today, 2017, Coverage of our neuroscience research
- “Consciousness is really under siege”Michael Pollan discusses his profile of my research, New Scientist, 2026
- If You Meet ET in Space, Kill HimGeorge Musser on cooperation and artificial empathy, Nautilus, 2024
- Linking sense of touch to facial movement inches robots toward ‘feeling’ painExpert commentary on responding to pain and experiencing it, Science News, 2020
- Do Humans Dream of Androids Dreaming?Human–AI relationships and research on machine feeling, USC Dornsife, 2020
- Castle in the CloudsAn early look at my work on how the senses come together, USC Dornsife, 2013
Talks & recordings
- Artificial feelings require embodied vulnerabilityASSC, Santiago, Poster with agent videos
- Homeostatic coupling for prosocial behaviorALife, Kyoto, Presented by my coauthor Naoto Yoshida
- Truth or consequences: Homeostatic self-regulation in artificial neural networksALife, Contributed talk
Background & appointments
I founded Feeling Machines in 2019. My research has ranged from neuroimaging and soft robotics to a research internship in deep learning at NASA’s Jet Propulsion Laboratory.
I remain an affiliate of USC’s Brain and Creativity Institute.
- PhD, Neuroscience, University of Southern California
- BS, Psychology, University of Michigan, Honors College
Positions & research resources
- 2019–nowFounder & Principal InvestigatorFeeling Machines
- 2021–2023Research Assistant Professor of PsychologyUniversity of Southern California
- 2014–2021Research Associate, then Senior Research AssociateUSC Brain and Creativity Institute
- 2019Research Intern, Deep Learning Technologies GroupNASA Jet Propulsion Laboratory
- Design of machines with feeling analoguesDamasio & Man, US Patent 12,118,449, 2024
Short bio & portrait
Kingson Man is a neuroscientist and AI builder studying the connections between life, feeling, and intelligence. His work at Feeling Machines investigates homeostasis, vulnerability, and adaptive behavior in artificial agents. He earned his PhD in neuroscience at USC. His research has appeared in Nature Machine Intelligence and Science Robotics, and is featured in Michael Pollan’s A World Appears and Valerio Jalongo’s documentary Wider than the Sky.
Clinical AI at City of Hope
Separately from Feeling Machines, I lead applied AI projects in oncology at City of Hope. I’m a Data Science Manager, with a courtesy appointment as Associate Professor at the Beckman Research Institute.
Research
At City of Hope, I lead work on HopeLLM, a clinical AI platform for extracting and retrieving information from oncology records. We evaluate model outputs with clinicians on real cases. Projects include clinical trial matching and RESiN, a method designed to examine every note before answering a clinical question.
Selected papers
-
Read Every Single Note (RESiN): Coverage-First Clinical Data Abstraction Beyond RAG
Armin Shayesteh Zadeh, Sakhawat Hossain Saimon, Sunmin Park & Kingson Man. NeurIPS GenAI4Health Workshop, Accepted demo, 2026.
A method designed to examine the full clinical note history before answering a question.
-
Enhancing Oncology-Specific Question Answering With Large Language Models Through Fine-Tuned Embeddings With Synthetic Data
Lu, Mehdinia, Man, Wong, Mao & Eftekhari. JCO Clinical Cancer Informatics, 2025.
Adapting embeddings to oncology records to improve retrieval for clinical question answering.
Earlier papers & demonstrations
- 2025End-to-end clinical trial matching on real-world unstructured clinical dataLu K, Mehdinia S, Mao A, Wong CW, Eftekhari Z, Man K. Machine Learning for Health (ML4H) 2025, demo track.
- 2024HopeLLM: A cancer specializing framework for multi-data abstraction from patients’ entire clinical note historiesLu K, Mehdinia S, Mao A, Bonner J, Pal S, Ebrahimi H, Li YR, Tam A, Wong CW, Eftekhari Z, Man K. Machine Learning for Health (ML4H) 2024, demo track.
- 2024Automated extraction of PSA values from unstructured medical records with a large language model for cancerEustace NJ, Man K, Lu T, Mehdinia S, Tam A, Ladbury CJ, Li YR. International Journal of Radiation Oncology, Biology, Physics 120(2, Supplement), e621-e622. Meeting abstract.
- 2024Use of artificial intelligence to derive International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) elements from unstructured medical recordsJaime-Casas S, Ebrahimi H, Man K, et al. The Oncologist 29(Suppl 1), S34-S35. Meeting abstract.
- 2023Clustering whole-exome sequences of breast cancer and associations with staging and molecular subtypeMan K, Lim A, Eftekhari Z, Wong CW. Journal of Clinical Oncology 41(16 suppl), e12563. Meeting abstract.
Press & talks
-
Meta, Company feature
Open Source AI is Leading to Breakthroughs in Healthcare, Education and Entrepreneurship
A conversation about our team’s work on clinical trial matching at City of Hope.
- Science Fiction or Real Science? AI + Health Hype vs RealitySXSW, Austin, A panel with Gabi Viljoen and Shawna Butler
Positions & resources
- 2021–nowApplied AI & Precision MedicineCity of Hope, Data Science Manager since 2025; previously Principal, Lead, and Senior Data Scientist
- 2023–nowAssociate Professor, courtesy appointmentBeckman Research Institute of City of Hope
- AI & machine learning tutorials for the California Teachers StudyOpen research notebooks, Team resource