What we — and AI — can learn from nature's intelligence

Episode Summary

Executive Summary: This TED Radio Hour episode explores “natural intelligence” in plants, animals, and the immune system, showing how living systems compute, count, adapt, and communicate. Through examples like the Venus flytrap, dragonflies, cytokines, and bioacoustics, the episode argues that intelligence is broader than human cognition and offers models for low-power AI, better health understanding, and conservation.

Main Topics: Venus flytraps and plant computation (Priority: 5/5): Greg Gage demonstrates that Venus flytraps can count touches and close only after repeated stimulation, reframing plants as active computational organisms rather than passive life forms. Mimosa, pea plants, and slime molds as intelligent systems (Priority: 5/5): The episode expands plant intelligence to touch-sensitive mimosas, decision-making pea plant experiments, and slime molds that process information without a brain, suggesting cognition can exist in non-animal forms. Redefining intelligence in biology (Priority: 5/5): Greg Gage argues intelligence is the ability to get what you want given what you have, implying that cells, plants, and other organisms all exhibit forms of intelligence through adaptation. Dragonflies and bio-inspired AI (Priority: 5/5): Francis Chance explains how dragonflies perform rapid interception calculations with very few neural steps, inspiring efforts to design low-power AI chips and more efficient navigation and defense systems. Immune system, inflammation, and behavior (Priority: 4/5): Keely Muscatel describes cytokines as immune messengers that drive sickness symptoms and also influence mood, social withdrawal, and craving for comfort, linking inflammation to mental and social behavior. Bioacoustics, AI, and interspecies communication (Priority: 5/5): Karen Bacher shows how machine learning helps decode animal and ecological soundscapes, from bats and whales to coral reefs, with applications in conservation, monitoring, and ethical questions about listening to other species.

Key Arguments: Natural systems often solve problems through computation, sensing, and adaptation rather than conscious thought. Venus flytraps count touches and time intervals to avoid wasting energy on false triggers. Plants and slime molds demonstrate that intelligence does not require a brain; cells themselves can process information. Dragonflies perform highly efficient interception calculations, making them a model for compact, low-power AI. The immune system not only fights infection but also shapes mood, motivation, and social behavior to promote recovery. Inflammation can be adaptive in the short term but harmful when chronic, contributing to stress spirals and disease. AI and bioacoustics can reveal hidden communication in nature and may support conservation, but they raise ethical concerns about eavesdropping on other species. Listening to natural soundscapes can help restore ecosystems, track endangered species, and reduce human impact on biodiversity.

Data Points: Venus flytrap trap reopening time: 24 to 48 hours - Greg Gage explains why the plant waits for repeated touch before closing, because reopening is slow and energy-intensive. Venus flytrap trigger threshold: 2 touches within about 20 seconds - The plant counts successive stimulations of trigger hairs before snapping shut. Dragonfly prey success rate: up to 95% - Francis Chance notes how effective dragonflies are at capturing chosen prey. Dragonfly response latency: about 50 milliseconds - Used to estimate the small number of neural computation steps available for interception. Single neuron processing time: about 10 milliseconds - Supports the estimate that dragonfly brains can use only a few sequential layers of computation. Estimated dragonfly computation layers: at most 4 layers - Chance says the hunting circuit can only have a few layers of neurons in sequence. Human brain power use: about 20 watts - Presented as a benchmark for low-power, brain-inspired computing. Neural network model accuracy for plants: about 70% - An algorithm trained on plant ultrasound could identify plant health states with this accuracy. Orca community size off the U.S. East Coast: less than 400 whales - Used to highlight the fragility of North Atlantic right whales and the importance of ship-avoidance systems. Ship-strike deaths in protected whale zone: 0 since launch - Bioacoustic real-time alerts have prevented whale deaths from ship strikes in that area. Right whale population reference: less than 400 - Karen Bacher describes the endangered population benefiting from ship-routing technology. Coral reef recovery method: healthy reef sounds - Scientists use soundscapes to help regenerate reefs, described as a form of music therapy for nature.

Pivotal Quotes: "The coolest thing is that the plant can count." — Greg Gage: A TED stage demonstration of Venus flytrap behavior and electrical signaling. "Being able to get to what you want given what you have." — Greg Gage: His concise definition of intelligence during the conversation about plants and slime molds. "Our commonality is greater than we knew." — Karen Bacher: Her closing argument that bioacoustics reveals deep connections across the tree of life.

Implications: Listeners are encouraged to broaden their definition of intelligence, see biology as a source of AI design, and recognize that health, conservation, and technology all benefit from learning how natural systems sense, compute, and communicate.