Dreaming Machines: Artificial Phronesis and the Future of Mind
Can a machine learn practical wisdom, the Aristotelian virtue of phronesis, or only imitate it? This paper starts from a question the author, Jacob Gino, asked at Full Sail University's March 2025 panel "How AI Is Impacting … Well, Everything." Drawing on the Ten Lens Ontology, it argues that today's large language models hold vast knowledge (episteme) and skill (techne) but lack phronesis. Phronesis is the grounded, experience-based judgment of what fits this person, here and now. Evidence that AI is sycophantic and morally inconsistent, including the Freysa prize contest, shows how persuadable these systems remain. For now, the paper proposes AI as a "wisdom mirror" that helps people reason better, drawing on inner-speech robot research. It then maps speculative avenues toward genuine artificial phronesis: wetware and living-neuron computing, a day-night robot that dreams by remixing its memories in a simulated world, priming-seeded search with checks against apophenia, and an AI lineage in which minds face finitude and pass knowledge to successors. Each avenue comes with testable conditions and named risks, including model collapse and machine suffering. The paper closes with a falsifiable benchmark: a wise machine's judgment must hold when a user pushes, flatters, or reframes the question.
Authors
- Vincent Baelin-Grayson (ORCID: https://orcid.org/0009-0008-7997-9448)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-16
- DOI
- https://doi.org/10.5281/zenodo.24288673
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- preprint