Maya (माया) describes the timeless tendency to mistaken representation for reality, attributing meaning or coherence to what only appears structured. Generative AI industrializes this tendency, producing fluent, persuasive outputs that simulate understanding without possessing it.
What was once an insight of Indian philosophy about perception has become a technological force that engineers appearance at scale. As systems grow more seamless, we risk confusing automation with cognition and probability with intention. The challenge before us is to carry an ancient awareness into a computational future designing AI that supports human discernment rather than deepening digital Maya.
A philosophical lens helps us see that the challenge with AI is not only technical, but perceptual. Maya describes a deeply human tendency: to project meaning, agency, and coherence onto convincing appearances. Generative AI does not create this tendency, it amplifies it at scale. By thinking philosophically, we move beyond feature design and ask deeper questions about belief, trust, authority, and discernment. That shift is essential if we want to design systems that support human sense-making rather than quietly distorting it.
In this talk Disha will lift the ontological veil that causes people to mistake the apparent for the real and elegantly transfer the Maya concept to AI:
| Philosophical Maya | AI Maya | |
| Effect | Minds project meaning onto chaos | LLMs generate fluent text without understanding |
| Result | Appearance of order | Appearance of intelligence |
| Danger | Confusing phenomenon with truth | Confusing fluency with cognition |
| Solution | Discernment (viveka) | Designing for sense-making |
