Envisioning an AI that embodies our shared humanity.
We are a team of interdisciplinary researchers dedicated to understanding the nature of what it means to apply good sense to artificial intelligence. We explore its societal impact and devise future scenarios where Good Sense in AI becomes an essential feature for human survival.
Core Premise
Common Sense ≠ Sound Judgement
We believe common sense does not necessarily equate to sound judgement. An AI system can hold encyclopedic common sense knowledge and faithfully reproduce consensus belief while still lacking sound judgement.
"Good sense"—a capacity for sound, context-sensitive practical and moral judgement—is conceptually distinct from the vague, commonly-held set of beliefs of a community. Our initiative focuses on how to foster genuine good sense in AI architectures.
The Three Pillars
Our Intellectual Foundation
Our framework synthesizes ethics, social ontology, and information philosophy to inform Good Sense for AI, by AI, to AI.
Normative ContentHume's corrected moral sentiment supplies the normative content of good sense. It relies on a "steady and general point of view" that provides a sentiment-based intersubjective standard, rather than optimizing for a fixed value-object.
Social MechanismEpstein's framework of anchoring and grounding supplies the social-ontological mechanism by which AI reshapes social reality. AI does not merely output facts; it re-anchors the frame principles of social kinds.
Stage & DynamicsFloridi's Fourth Revolution supplies the ontological stage—the infosphere—where living "onlife" occurs. It provides the recursive society-technology feedback loop that can become virtuous or vicious depending on our moral standpoint and the consequences of AI.