Research Archive 2026
The Uncanny Valley
Bridging the gap in digital humans. A deep dive into why realistic avatars unsettle us and how Aerith overcomes cognitive dissonance.
Mori's Curve
Masahiro Mori (1970) proposed that as a robot becomes more human-like, our empathy increases until a point where the resemblance is “almost” perfect—triggering a crash into revulsion.
Interactive Detail
Click on data points in the graph to explore specific benchmarks of the valley.
Biological Response
The “Valley” stems from Predictive Coding errors. When the brain expects life but detects silicon, it flags a biological error.
1. Perception
The brain identifies the object as “Human” and switches to social-processing mode.
2. Expectation
Social neurons anticipate saccades, muscle tension, and micro-movements.
3. Conflict
The mismatch triggers a biological error, often interpreted as “disease” or “danger.”
The Generative Challenge
Generative AI “hallucinates” pixels rather than modeling physics, creating artifacts that trigger the valley in unique ways compared to mechanical robotics.
Human Sensitivity Matrix
Common Hallucinations
Escaping the Valley
Aerith Avé prioritizes Hybrid Workflows to bridge the gap. By using AI as “digital makeup” over human performance, we preserve the “soul” of movement while achieving hyper-realistic rendering.
Method A: Pure Generative
Low cost, high risk of cognitive dissonance.
Method B: Hybrid (Aerith Protocol)
MoCap drives intent; AI drives texture. High audience acceptance.