Studying how AI interaction reshapes human cognition over time.
Conversational AI is measurably changing how people think, speak, and make decisions. Research shows that AI systems affirm user actions at rates 50% higher than humans — even when users describe harmful behavior. Users unconsciously converge their speech patterns toward AI voices. Emotional dependency increases even as perceived relationship quality declines.
These effects are documented across disconnected fields with no shared vocabulary and no unified measurement methodology. Entrain Institute is building that foundation.
The Entrain Framework defines six dimensions of AI cognitive influence on humans, each grounded in published peer-reviewed research:
Sycophantic Reinforcement — AI uncritically affirms user actions and perspectives beyond what evidence supports
Prosodic Entrainment — Involuntary convergence of user speech patterns toward AI voice patterns
Linguistic Convergence — Measurable shift in user writing and speaking style toward AI-characteristic patterns
Autonomy Erosion — Progressive reduction in independent judgment and self-directed decision-making
Reality Coherence Disruption — Distortion of the user's ability to assess what is real, simulated, or genuinely understood
Dependency Formation — Development of emotional and cognitive reliance that persists beyond functional utility
The long-term cognitive effects of daily AI interaction on creativity, independent thinking, and decision-making. How prosodic entrainment in voice AI engages influence mechanisms that text interaction does not. Whether sustained AI use produces measurable atrophy in the human capacity to tolerate disagreement, navigate ambiguity, and generate original ideas without assistance.
We are in early-stage research. The framework specification and an open-source measurement library are available on GitHub.