TED Radio Hour: "Decoding Nature’s Hidden Patterns"

Host: Manoush Zomorodi (NPR)
Guests: Sarah Beery (MIT AI & Conservation), Jeff Reed (Computational Linguist, Grizzly Systems)
Date: November 28, 2025

Overview: Decoding Nature’s Language

In this episode, TED Radio Hour dives into how technology is revolutionizing our understanding of nature’s hidden patterns. Host Manoush Zomorodi explores how AI and emerging tools enable scientists and everyday people to gather, process, and interpret vast ecological data—from cataloguing species to decoding wolf howls. The show emphasizes the blend of optimism and realism required to harness tech for conservation, underscoring that the future of protecting wildness depends on everyone pitching in—humans and machines alike.

Key Discussion Points and Insights

1. The Limits of Our Knowledge and the Biodiversity Crisis

• Humanity’s Gaps in Understanding Nature

  • Despite centuries of study since Darwin, humanity has only documented about 2 million species, while expert estimates suggest the planet harbors between 10 and 100 million (01:21–02:01).
  • “If you want to protect a species, it’s not enough to just know that it exists… you have to understand what it needs, what it eats, how it relates to its environment.” —Sarah Beery (02:18)

• Ecosystem Losses

  • Over 70% of wildlife has been lost since 1970—an “alarming rate” that’s “shocking,” as Sarah puts it (02:49).

2. Citizen Science & Data Explosion

• The Rise of Community Science

  • Platforms like iNaturalist and eBird allow anyone to collect and contribute crucial biodiversity data via smartphones (03:15–03:39).
  • Sarah notes, “Community science data makes up probably 90% of all biodiversity data we have and have ever collected” (03:51).

• Unlocking the Hidden Layers

  • Every field photo contains a wealth of info—species behavior, habitat, environmental conditions—not just species ID (05:39–06:21).

• From Data Avalanche to Insights

  • With 200 million images on iNaturalist, “we’re sitting on an ecological goldmine… the problem is accessing the knowledge efficiently” (TED Stage, 07:50; 07:58).

3. AI as a Transformative Tool for Ecology

• The Challenge of Scale

  • “Assuming it takes you about a second to look at every image, you would need to work full time for 40 years to look through all the images in iNaturalist alone. And this is where AI is transformative.” —Sarah Beery (TED Stage, 07:58)

• AI That Answers Naturalists’ Questions

  • Sarah’s team created “Enquire,” an AI system that lets ecologists query massive datasets using natural language—no coding required (09:54).
  • Example: Researchers studied forest regrowth post-fire by searching images for evidence of burns and regrowth. The findings? Severely burned areas regrow with more deciduous trees, while less severely burned areas favor conifers (09:54–11:18).

  “Understanding the implications of the way that climate is changing… helps us understand what to prioritize, where to prioritize, how to take action to protect species.” —Sarah Beery (11:28)

• Real-Time Conservation Impact

  • Data-driven adaptive management: e.g., cities turning off lights during major bird migrations to prevent deadly window strikes. “So easy, so easy. But we need to know when and where it’s happening to be able to take that step.” —Sarah Beery (12:00–13:04)

4. AI’s Environmental Footprint & Collaboration

• Tech Optimism Meets Pragmatism

  • Manoush raises the concern: “All of the energy that these data centers require are just making it worse. How do you see it?” (16:21)
  • Sarah clarifies: Field AI for ecology is designed to be lightweight to run in remote, resource-scarce environments—far less carbon-intensive than popular generative AI (16:21–17:42).

• The Importance of Fieldwork

  • Real impact comes from co-designing tech with conservationists on the ground. Sarah’s salmon escapement project uses sonar and AI to monitor salmon recovery and river ecology post-dam removal—opening 800 square miles of new habitat (17:56–22:03).

5. Balancing Tech-Saviorism and Real-World Impact

• AI as Aid, Not Savior
  • Sarah cautions against “tech saviorism”: “AI isn’t a magic wand… it can be a really amazing tool… but I really think it’s vital that we have these experts very tightly connected in the loop” (22:54–24:27).
  • The future of conservation lies in “our ecological databases, both the ones we have now, but also the ones we have yet to collect” (TED Stage, 25:06).

6. Decoding the Language of Wolves

The Cry Wolf Project

• History and Motivation

  • Jeff Reed, with a background in tech and linguistics, returns to Yellowstone after wolf reintroduction and embarks on a mission to decode their howls (28:25–29:15).

• Building New Bioacoustic Tools

  • Outdated equipment led Jeff to invent robust acoustic recorders capable of surviving Yellowstone's extremes and collecting terabytes of audio (30:01–31:18).

• Surplus Data: The AI Solution

  • “We’ve recorded over 200,000 hours, which would take me roughly 50 years to listen to.” —Jeff Reed (31:18–31:40)
  • AI tools now sift through this trove, finding and classifying howls and uncovering patterns even humans might miss (37:32–38:26).

  “There are clustering tools… that allow us to use software to say, what are we missing as humans? What might be another way to look at this howl?” —Jeff Reed (38:06)

• How Wolves Communicate

  • Wolves initiate chorus howls led by dominant pack members—AI helps identify who decides to chorus and may even help count wolves via audio alone (39:48–41:09).
  • Wolves combine barks and howls into “sentence-like structures”—the bark as alarm, the howl as a location beacon (41:09–42:46).
  • Case study: Wolf 907’s unique howl allowed packmates to identify her; context was crucial for interpreting meaning (43:10–47:51).
    • “[When] she howled like this for half an hour,” another wolf eventually howled back—keeping track of the group (47:52).

Memorable Quotes & Moments

• Sarah Beery (on hope in conservation):

  “We stand at a unique point in history. We have both an unprecedented biodiversity crisis, but we also have unprecedented tools to address it.” (TED Stage, 24:27)

• Sarah Beery (on participation):

  “Everyone can contribute, everyone can collect data and upload it to platforms like iNaturalist. Every photo uploaded, every sound recorded… is a piece of the puzzle.” (TED Stage, 25:09)

• Jeff Reed (on the language of wolves):

  “I don’t foresee a Google Translate for wolffish. Right. Because what you’re really trying to do is get in the mind of a wolf, and that’s different than just transcribing their different types of howls…” (48:46)

• Jeff Reed (on what tech can and can’t do):

  “You tend to protect what you love, and you only love what you understand.” (49:33)

• Jeff Reed (confronting loss of wildness):

  “If your body represented the total weight of all the world's land mammals… your right forearm would be what's left of the wild ones. The rest… is us, our livestock, and our pets.” (TED Stage, 50:54)

Key Segments and Timestamps

| Timestamp | Segment/Topic | Speaker(s) | |-----------|--------------------------------------|----------------------| | 00:44–02:01 | How little we know about biodiversity | Sarah Beery | | 02:18–02:49 | The challenge of preserving unknown species | Sarah Beery | | 03:39–04:24 | Citizen science & iNaturalist’s role | Sarah Beery | | 05:21–06:21 | Hidden info in field images | Sarah Beery | | 07:50–08:13 | AI for scaling biodiversity insight | Sarah Beery (TED) | | 09:54–11:18 | Enquire AI for ecological queries | S. Beery, M. Zomorodi| | 12:00–13:04 | Bird migrations, city lights & tech aid | S. Beery | | 16:21–17:42 | AI’s environmental footprint | S. Beery, M. Zomorodi| | 17:56–22:03 | Fieldwork, dam removal, salmon monitoring | S. Beery | | 22:54–24:27 | Dangers of tech saviorism | Sarah Beery | | 25:06–25:40 | Everyone’s role in data collection | Sarah Beery (TED) | | 26:16–29:15 | Wolf howls, history & symbolism | Jeff Reed | | 30:01–31:40 | Engineering new acoustic technology | Jeff Reed | | 37:32–38:26 | AI for analyzing howls | Jeff Reed | | 39:48–41:09 | Chorus howls, pack behavior, counting wolves | Jeff Reed | | 41:09–42:46 | Combinatorial “language” in wolf calls| Jeff Reed | | 43:10–47:51 | The story of Wolf 907 | Jeff Reed, M. Zomorodi| | 48:46–49:33 | Limits and promise of AI in animal language | Jeff Reed | | 50:54–51:39 | On the future of wildness | Jeff Reed (TED) |

Tone & Style

The conversation blends wonder at ecological complexity, grounded analysis of technological solutions, and a respectful skepticism of AI hype. Both main guests (Sarah Beery and Jeff Reed) are practical idealists: passionate about tech’s ability to unlock nature’s secrets, but wary of overstating its power or losing sight of what conservation really demands—active, informed participation from many.

Conclusion

This episode provides a thought-provoking look at the intersection of artificial intelligence, citizen science, and old-school observation. It echoes a call to action: we need all hands—human, machine, amateur, and expert—to decode, understand, and protect the living world before more of its patterns slip into silence.