Project CETI (Cetacean Translation Initiative) used AI to decode sperm whale vocalizations, identifying a “phonetic alphabet” of clicks (codas) with syntax akin to human language.
Technical Insight:
- Deep learning models analyzed 100,000+ whale recordings from hydrophones.
- Discovered contextual “dialects” between pods, suggesting cultural transmission.
Applications:
- Conservation: Monitoring whale stress from ship noise.
- Evolutionary Biology: Insights into language origins.
Criticism: Some linguists argue anthropomorphizing animal communication risks flawed conclusions.