John's Research
My research focuses on issues in the philosophy of cognitive science, as well as issues in the ethics of AI, in particular I am interested in both philosophical and scientific problems surrounding self-knowledge and metacognition and the possibility of ethical technology. I have published on ethical AI, metacognition, embodiment, the phenomenology of reasoning, and on the foundations of epistemic agency, as well as some vulnerabilities this agency has in the digital age.
Talk: Why AI Cannot be Trustworthy (Yet)
Talk given at Dresden University (Sep, 2023)
Abstract: Many current policies and ethical guidelines recommend developing “trustworthy AI”. Here we argue that developing trustworthy AI is not only unethical, as it promotes trust in an entity that cannot be trustworthy, but it is also unnecessary for optimal calibration. Instead, we show that reliability, exclusive of trust, entails the appropriate normative constraints that enable optimal calibration and mitigate the problem of vulnerability that arises in high-stakes hybrid decision-making environments, without also demanding, as trust would, the anthropomorphization of artificial assistance and thus epistemically dubious behavior. Here, the normative demands of reliability for interagential action are argued to be met by an analogue to procedural metacognitive competence (i.e., the ability to evaluate the quality of one’s own informational states to regulate subsequent cognitive action). Drawing on recent empirical findings that suggest providing precision scores (such as the F1-score) to human decision-makers improves calibration on the AI-system, we argue that precision scores provide a good index of competence and enables humans to determine how much they wish to rely on the system
Publications
Dorsch, J., & Deroy, O. (2024). The impact of labeling automotive AI as "trustworthy" or "reliable" on user evaluation and technology acceptance. arXiv:2408.10905. https://doi.org/10.48550/arXiv.2408.10905
Dorsch, J. and Deroy, O. (2024). Quasi-Metacognitive Machines: Why We Don’t Need Morally Trustworthy AI and Communicating Reliability is Enough. Philosophy & Technology, 37, 62. https://doi.org/10.1007/s13347-024-00752-w
Dorsch, J. (2024). Can You See a Ganzfeld? Critical Notice of Schellenberg’s Unity of Perception. International Journal of Philosophical Studies. https://doi.org/10.1080/09672559.2024.2360782
Dorsch, J. (forthcoming). Why AI Cannot Be a Bona Fide Moral Agent (Yet): Trustworthiness, Moral Normativity, and Giving a Damn. American Philosophical Quarterly.
Dorsch, J & Moll, M. (forthcoming). Explainable AI for Decision Support Systems: The Theory of Human-AI Epistemic Quasi-Partnerships. Müller, V. C. (Ed.) Philosophy of Artificial Intelligence: The State of the Art 2023. Springer.
Dorsch, J. (forthcoming). Embodied Metacognition: The Role of the Body on the Path to Self-Knowledge. In book series "Studies in Brain and Mind", Springer Nature.
Dorsch, J. (2023). Are Noetic Feelings Embodied? The Case for Embodied Metacognition. Philosophical Psychology. https://doi.org/10.1080/09515089.2023.2197937
Dorsch, J. (2022). Hijacking Epistemic Agency: How Emerging Technologies Threaten our Wellbeing as Knowers. Proceedings of the AAAI and ACM on Artificial Intelligence, Ethics, and Society 5. https://doi.org/10.1145/3514094.3539537
De Preester, H. & Dorsch, J. (2021). Descartes’s Passions of the Soul, Epistemic Affect and the Challenges for Interoception Research in Emotions. Danish Yearbook of Philosophy, 54(1), 65-92. https://doi.org/10.1163/24689300-bja10021
Dorsch, J. (2018). Prescribing Proficiency: Ethical Medical Practice in English-Speaking Contexts. German American Institute, Tübingen. (Course book for teaching cultural competency and ethical medical practice in patient encounters, published by the German American Institute and actively used in the curriculum).
Dorsch, J. (2017). On Experiencing Meaning: Irreducible Cognitive Phenomenology and Sinewave Speech. Phenomenology and Mind 12, 218-227. https://doi.org/10.13128/PHE_MI-21120
Dorsch, J. (2016). Irreducible Cognitive Phenomenology and the 'Aha!' Experience. Phenomenology and Mind 10, 108-121. https://doi.org/10.13128/PHE_MI-20095
Presentations (peer-reviewed)
Dorsch, J. (2024). Algorithm Aversion: Trustworthy vs Reliable AI. Preliminary Results from a Between-Subject Study on Technology Acceptance. Bavarian Research Institute for Digital Transformation. Bavarian Academy of the Sciences and Humanities, Munich.
Dorsch, J. (2024). A Puzzle for Explainable and Ethical AI: Avoiding Algorithm Aversion and Cultivating Epistemically Virtuous Decision-Making. Cognition, Values & Behaviour Research Group, LMU, Munich.
Dorsch, J. (2024). A Puzzle for Explainable and Ethical AI: Avoiding Algorithm Aversion and Cultivating Epistemically Virtuous Decision-Making. Ethics of AI Research Colloquium, LMU, Munich.
Dorsch, J. (2023). Explainable AI for Decision Support Systems: The Theory of Epistemic Partnerships. Society for the Philosophy of AI, 5. University of Erlangen.
Dorsch, J. (2023). Why AI Cannot be Trustworthy (Yet). Workshop on Human-AI Interactions: Peace Research Institute, Oslo and Cognition, Values & Behaviour Research Group, München.
Dorsch, J. (2023). Why AI Cannot be Trustworthy (Yet). NATO: MSCoE Group (North Atlantic Treaty Organization: Modelling & Simulation Centre of Excellence), München.
Dorsch, J. (2023). Why AI Cannot be Trustworthy (Yet). Cognition, Values & Behaviour Research Group, Ludwig-Maximilians-Universität, München.
Dorsch, J. (2023). Why AI Cannot be Trustworthy (Yet). socialBRIDGES: Trust: Bridge Between Society and Technology. Universität Dresden.
Dorsch, J. (2023). Why AI Cannot Be Trustworthy Even If We Needed It to Be. Cognition, Values & Behaviour Research Group, Ludwig-Maximilians-Universität, München.
Dorsch, J. (2023). EU Regulations of AI - What They Mean for Philosophers and Friends of Philosophers Working in AI. Cognition, Values & Behaviour Research Group, Ludwig-Maximilians-Universität, München.
Dorsch, J. (2023). How To ChatGPT - Ethical Digital Assistance and Effective Prompt Engineering. Cognition, Values & Behaviour Research Group, Ludwig-Maximilians-Universität, München.
Dorsch, J. (2023). Knowing Ourselves Together Through Embodied Entanglement: The Role of Embodied Confidence on the Path of Self-Knowledge. 4th International Conference on Philosophy of Mind: 4E’s Approach to Mind/Brain. Universidade Católica Portuguesa. Braga, Portugal
Dorsch, J. (2023). Co-Learn: Trust in AI Assistants. Bavarian Research Institute for Digital Transformation. Bavarian Academy of the Sciences and Humanities, Munich.
Dorsch, J. & Ripp, I. (2022) Co-Learn: Digital Workplace and Human-AI Assisted Training. Bavarian Research Institute for Digital Transformation. Bavarian Academy of the Sciences and Humanities, Munich.
Ripp, I. & Dorsch, J. Co-Learn: Humans in the Learning Loop. Bavarian Research Institute for Digital Transformation. Bavarian Academy of the Sciences and Humanities, Munich.
Dorsch, J. (2022). Hijacking Epistemic Agency: How Emerging Technologies Threaten Our Wellbeing as Knowers. The Fifth Association for the Advancement of Artificial Intelligence & Association for Computing Machinery Conference on Artificial Intelligence, Ethics, and Society, Keble College, University of Oxford.
Dorsch, J. (2022). What Do You Feel When You Feel Confident? The Bayesian Brain and Embodied Metacognition, Consciousness Club, Wellcome Centre, University College London, (online).
Dorsch, J. (2021). Embodied Metacognition: How We Feel Our Hearts to Know Our Minds, Cognition, Values & Behaviour Research Group, Ludwig-Maximilians-Universität, München, (online).
Dorsch. J. (2021). The Case for Embodied Metacognition: Between Evaluative and Metarepresentational Accounts, Metacognition: New Directions Conference, University of London, (online).
Dorsch, J. (2021). Are Noetic Feelings Associated with Semantic Memory Embodied? Centre for Philosophy of Memory, Université Grenoble Alpes, (online).
Dorsch, J. (2020). In Defense of the Feeling of Knowing: Tracking the Target of the Tip-of-the-Tongue Experience. Mental Actions Workshop, University of Edinburgh.
Dorsch, J. (2019). What's on the Tip of Everybody's Tongue: Predictive Processing and the Tip-of-the-Tongue Experience”. Mind and Cognition Seminar, University of Edinburgh.
Dorsch, J. (2018). The Phenomenology of Meaning: Constitutive Cognitive Phenomenology. Philosophy of Neuroscience Group Research Forum, University of Tübingen.
Dorsch, J. (2017). Perception: a discriminating capacity or a predictive capacity? Presented at the 9th Annual Junior Neuroscientist Conference, University of Tübingen.
Dorsch, J. (2016). Irreducible Cognitive Phenomenology and the 'Aha!' Experience. Philosophy of Neuroscience Group Research Forum, University of Tübingen.