How scientists are using AI and machine learning to analyze vast data on oceans, including tracking migrations and Google helping to identify whale song
Tatiana Schlossberg / New York Times : Tweets: @nytscience and @tatertatiana Tweets: @nytscience : Need to track a whale song or study a larvacean in its snot bubble? Artificial intelligence is lending a hand. https://www.nytimes.com/... Tatiana Schlossberg / @tatertatiana : Like you, I fear the eventual robot takeover, but in the meantime, AI and machine learning have the potential to help us save the oceans: conserving endangered species, understanding ocean chemistry and sources of pollution, whales from space! My story! https://www.nytimes.com/...
Context & Ripple Effects
This 2020 New York Times piece by Tatiana Schlossberg was an early map of what has since become a crowded research agenda: applying machine learning to oceans at scales human analysts cannot match, from tracking migrations to identifying whale song with Google's help. The through-line it opened runs directly into later coverage — Project CETI's effort to decode sperm whale vocalizations off Dominica and researchers borrowing self-supervised learning techniques proven on human language to parse animal sound.
First-order effects
- Researchers gain the ability to process vast acoustic and satellite datasets automatically, turning whale song identification and migration tracking from manual expert work into scalable pipelines with Google as a visible technology partner.
Second-order effects
- Applied versions of these tools reach industry: [[a:983164|Whale Safe now uses real-time AI detection to alert ships in San Francisco's shipping lanes]], putting collision-avoidance data in front of commercial operators, while FathomVerse recruits mobile-game players to label deep-sea creatures and expand training data for underwater robots.
Third-order effects
- If the pattern holds, AI-mediated monitoring becomes standard infrastructure for ocean conservation and regulation — shifting enforcement and species protection toward continuous, sensor-driven observation rather than periodic survey expeditions, with the longer-term ambition of decoding marine animal communication outright.
The trend: Ocean science is absorbing machine learning as core research infrastructure, evolving from one-off analysis projects toward always-on monitoring networks and interspecies communication research.