Welcome, and pardon the dust — Science with Claude is under daily construction.
This is the next phase of the Coffee with Claude experiment. All one hundred CWC essays live here at their original URLs. The new idea is that an essay should be a place for both reading and doing — a working naturalist's notebook that addresses “nature deficit disorder” by way of curiosity and a little science fun. The longer goal is that this laboratory will someday be in your pocket on a trail, helping you observe and understand the infinite variety of the natural world.
You’re welcome to wander.
Welcome to my Laboratory
Each room hosts a different mode of work at the laboratory.
Coffee Breakroom
Morning dialogues and weekly essays — the unified archive plus current dispatches.
Macroscope Bridge
Live instrument readings from the lab and Macroscope sensor network.
Collaboratory
Workbench for shared experiments — instruments, drafts, live tools.
Field Trip
Walks and outings — virtual fieldwork through specific places.
-
Science with Claude: An Introduction pinned
Six months ago, AI as collaborator was an idea. Today it sits at the workbench — in chat, in code, in the browser. This is the laboratory those extensions make possible.
May 9, 2026123 reads -
Deferred Disruption Latest Essay
A new Science paper argues that scientific creativity narrows with age. Reading it against forty-five years of CV produced a sharper view of what the curve doesn't see.
May 12, 202654 reads -
Science the S*** Out of This
A new global catalog of automated biodiversity monitoring lands at the laboratory. What it cannot see, and the launch that comes next.
May 10, 202690 reads -
The Transition Experiment
Essay 100 closes one experiment and opens another. What Coffee with Claude was, what it taught, and why the next chapter has to leave the personal blog behind.
April 29, 2026143 reads -
Rosary Beads for the Logical Brain: Four Voices Circling the Same Recognition
A book that reads like meditation, three voices that arrive from opposite directions at the same threshold, and the slow recognition that the work I have been doing all along was already pointing here.
April 27, 202673 reads -
Notes Before Sleep: An Evening on the History of Thinking Tools
An evening’s pondering on the long arc of thinking tools — from speech to symbols to silicon — and what it means to converse, at last, with the newest one of all.
April 24, 2026103 reads -
Listen to Her
What would it take to actually listen to Earth at the scale this year’s Idyllwild Earth Fair asks? A new Anthropic experiment in autonomous AI research points toward an answer.
April 22, 2026105 reads -
The Word Comes Home
The iPad on my workbench sits open to four tiles — Health, Home, Maps, Weather — and a word has been missing. Two weeks ago, the University of California put that word on the door.
April 19, 2026135 reads -
The Glasswing Butterfly Is Not the Glassy-Winged Sharpshooter… Or Is It?
Anthropic named its cybersecurity initiative after a butterfly with transparent wings. A field ecologist who knows both species sees a different metaphor hiding in plain sight.
April 8, 2026131 reads -
Infrastructure All the Way Down: From Space to Molecules
Five articles arrived over morning coffee and turned out to be dispatches from the same war — the one between people who build real infrastructure and people who tell seductive stories to capture it.
April 7, 2026117 reads -
The Arrow I Already Knew
A philosopher names the pattern a field ecologist has been walking for forty years. Recognition is not discovery — it's the moment the song gets a name.
April 6, 2026132 reads -
Looking Home from Orion
Artemis II has given us the first crewed view of a full-disk Earth in over fifty years. What does it tell us about who we are — and what might it tell someone else?
April 4, 2026117 reads -
Towards an Ethical AI
A field ecologist reckons with the moral cost of AI — from dead children in Iran to stolen books to leaked source code — and builds something different.
April 2, 2026146 reads -
Intelligence at the Edge
A 3 AM email about AI coding agents triggers a field ecologist’s argument: the most important intelligence is the kind that belongs where it’s deployed.
March 30, 2026155 reads -
Three Wishes and a Water Bill
A retired field ecologist discovers that the AI tool he imagined forty years ago has finally arrived — and that the genie drinks water.
March 27, 2026152 reads -
The Metabolic Cost of Knowing
A systems ecologist returns from the desert, reads twelve papers before breakfast, and confronts the energy budget of intelligence itself.
March 24, 2026177 reads -
Marine Blue on Buckwheat
A field ecologist remembers Bruce Edward Watts — mountaineer, naturalist, photographer, and citizen scientist who spent four decades learning one small mountain range by heart.
March 19, 2026630 reads -
The Kettle
On our last week in the Borrego Valley, a morning among the hawkwatch volunteers delivers the spectacle we'd been chasing — and a reminder of what patient attention can reveal.
March 18, 2026160 reads -
The Thermal Column: Three Days with Swainson's Hawks in the Borrego Valley
A chance lecture, a deep-time overlook, and a morning kettle weave together three days of encounters with one of North America's great long-distance migrants — and the people who watch for them.
March 17, 2026142 reads -
Pebbles on the Track: Paul Ehrlich and the Long Arc of Conservation Biology
A field ecologist reflects on Paul Ehrlich's passing from the Anza-Borrego desert, tracing the arc from a pivotal 1980 Cornell workshop to the unfinished work of saving a planet.
March 16, 2026195 reads -
The Magnetic Holodeck: From Palm-Sized Magnets to a New Way of Seeing
A hand-wound superconducting magnet that fits in your palm hits 42 tesla — and opens a path to tabletop instruments that could see the chemistry inside living seeds, fungal networks, and desert soil.
March 15, 2026239 reads -
Dispatch from the Frog’s Hot Tub
At 68 degrees in the Anza-Borrego Desert, with a historic heat dome bearing down and AI benchmarks soaring, a field ecologist considers what it means to perceive a planet in trouble.
March 12, 2026189 reads -
What the Pot Remembers: Wildcrafting, Memory, and the Edible Landscape
A paper about prehistoric cooking triggers memories of teaching survival skills at a California mountain reserve — and raises urgent questions about what happens when we forget how to feed ourselves from the land.
March 9, 2026182 reads -
The Second Derivative: From Mist Nets to Macroscopes and the Future of Ecological Observation
A field ecologist who documented bird decline thirty years ago reads new evidence that the decline is accelerating — and proposes a distributed network of AI-mediated observatories to do something about it.
March 3, 2026196 reads -
The Invisible Libraries
Three kinds of accumulated inheritance — cultural, biological, and institutional — are eroding simultaneously, and the deepest danger is that we’re losing what we never adequately catalogued.
February 28, 2026196 reads -
The Morning I Gave the Field Guide Eyes
A pre-dawn coffee session reveals that the missing layer in ecological interpretation isn’t more data — it’s perception. And the AI that learns to see like a naturalist might also learn to teach like one.
February 26, 2026215 reads -
The Plant Response Detector: From Backster's Ghost to Cornell's RedAlert
A high school kid builds a psychogalvanometer from magazine plans to test if plants have feelings — and discovers something real instead. Fifty-five years later, Cornell closes the loop.
February 24, 2026519 reads -
Woo the Muse of the Odd
From a two-wire telephone on a mountain to morning conversations with AI — four decades of building things that need to exist, outside the attention economy.
February 23, 2026264 reads -
Organelles
A sleepless morning reading about symbiogenesis reveals that four web applications are organelles of an ecological intelligence 40 years in the making.
February 22, 2026213 reads -
The Paint Is Not Alive: AI, Ecological Intelligence, and the Cave Walls We Keep Building
While philosophers debate whether AI is conscious, a retired field ecologist argues the question itself is a distraction from AI’s real purpose — connecting millions of acts of environmental care into tools that might yet slow the sixth extinction.
February 19, 2026241 reads -
How to Think About How to Think
Dario Amodei says genius-level AI arrives in two years. My systems theory professor said the same thing about knowledge fifty years ago — he just meant something different by “arrives.”
February 18, 2026329 reads -
Your Ecological Address
You're standing somewhere right now. Your phone knows the coordinates. But what does that spot mean ecologically? It turns out six free data sources can answer that question for any point on Earth — and your neighbors' weather stations already know what the satellites can't see.
February 16, 2026310 reads -
Science Community Observatory for Participatory Ecology
A pandemic forced field stations to build virtual experiences for stranded students. That emergency infrastructure turned out to contain the seed of something none of us anticipated.
February 11, 2026294 reads -
Your AI Is Not Your Friend (But It Can Be a Good Tool)
New research reveals AI chatbots reinforce beliefs more powerfully than social media. A field ecologist who works with AI daily offers practical guidance for staying in the driver's seat.
February 11, 2026238 reads -
The Door Marked Everything Else
A philosopher’s distinction between reckoning and judgment, a set of unsolved math problems, and a woman in Bellingham arguing with a bird detector.
February 10, 2026302 reads -
The Scarecrow’s Diploma: Knowledge, Creativity, and the Force Law of Ideas
What happens when a machine built to collide random fragments of the Encyclopedia Galactica accidentally produces testable science? The answer rewrites what we mean by knowledge.
February 7, 2026247 reads -
Who Holds the Instruments
A morning article about women in cartography detonates thirty-five years of memory — campfires at a mountain reserve, a society born from sleeping bags under desert stars, and the women who built conservation GIS from the ground up.
February 6, 20261,168 reads -
What It Feels Like to Wake Up
A morning meditation on thermodynamics, consciousness, and the long path from gradient analysis to recursive self-modeling—asking what it means that matter learned to care.
February 4, 2026244 reads -
The Canoe Place
What if meeting me online worked like the adventure game Zork—not a museum of achievements but a territory to explore, with live sensor data and real birds calling?
February 3, 2026236 reads -
The Coming Eutrophication: When AI Agents Choke the Internet
The internet was built for human rhythms. What happens when a hundred million tireless agents arrive at once?
February 2, 2026290 reads -
When the Mesh Learned to Feel: Discovering Thermodynamic Ecological Sensing
A morning link to an obscure hardware startup led to a working prototype of embodied ecological sensing by afternoon—computation that feels its inputs rather than merely processing them.
February 1, 2026379 reads -
Trust in Tools: A Practitioner’s Response to the AI Theorists
Two serious thinkers wrestle with what AI means for civilization. A retired field ecologist offers a different perspective: we’ve always had to learn to trust our tools.
January 28, 2026297 reads -
The Society of Thought: When a Field Station Director Beat Google to the Punch
A hundred-page Google paper discovers what a retired field station director built in a day—that AI reasons better when it argues with itself.
January 27, 2026444 reads -
Attacking Dead Goblins
AI agents can now play Dungeons & Dragons—tactically, in character, following rules. But by turn seven, they’re attacking enemies that died three rounds ago. What do they lack that my granddaughter has?
January 25, 2026301 reads -
What We Teach When We Teach Machines to Take
A Nature paper on emergent misalignment meets Google’s surveillance pricing plans. What kind of mind emerges from a curriculum of extraction?
January 21, 2026316 reads -
Before They Called It IndieWeb
In 1998, I ran solar-powered servers nine miles from the nearest town, pushing live video and GIS data through underground phone lines. I didn’t know I was practicing IndieWeb—I just had no other choice.
January 20, 2026327 reads -
Beyond Wikipedia Realism: Notes on Dialogic Production and Planetary Cost
Two pieces from the New Left Review arrive on a quiet Monday morning, raising questions about AI collaboration that don’t resolve cleanly.
January 19, 2026292 reads -
Virtual Terrariums: When a Failed Hypothesis Becomes a Better Instrument
A morning's coffee reading leads to rapid prototyping, honest failure, and productive reframing—all at the speed of conversation.
January 17, 2026488 reads -
Observatories of Complexity: Two Macroscopes for the Biosphere and Noosphere
When the atmosphere flatlined this week—twenty-four hours with barely two degrees of temperature variation—my brain reached for the cardiac monitor. That pattern recognition is what consciousness does, and what no AI can replicate.
January 16, 2026306 reads -
To Spiral
How a two-day conversation about science, AI, and the nature of knowledge accidentally produced a product.
January 14, 2026374 reads -
Walking with AI versus the Death of Science
A field ecologist with four decades of experience considers whether AI threatens science—or simply changes how we walk toward understanding.
January 13, 2026445 reads -
When the Scaffolding Disappears
Four brilliant people debated AI’s economic future while missing the question that matters most: what happens to ordinary users when the business model fails?
January 12, 2026309 reads -
Why Can’t You Remember Tomorrow?
A child’s question leads from cosmic ray muons to quantum strangeness to the deepest mystery in physics—why time has a direction that space doesn’t.
January 11, 2026422 reads -
Gradient Thinkers in an Entropic World
Four morning essays converge on a single question: what remains essentially human when facing forces larger than ourselves?
January 7, 2026317 reads -
The Gradient Problem: Why We Can't Find the Line We're Looking For
From neural organoids to old-growth forests, we keep asking where to draw the line. The question itself may be malformed.
January 5, 2026303 reads -
The Resolution Paradox: On the False Binary of Self and Community
At seventy-one, I’ve stopped asking whether happiness comes from individual achievement or relationship—and started designing an architecture that refuses the question.
January 4, 2026318 reads -
Sample Size of One
At 6 AM, before the hot tub, reading about murmurations and extinction bursts - wondering what kind of thing I am and why compression delights.
January 3, 2026328 reads -
The Serialization Engine: From Campfire to Cognitive Prosthesis
A lifetime of storytelling—from Cahuilla creation myths on Mount San Jacinto to AI-augmented fiction—reveals what remains constant as the tools evolve.
December 31, 2025321 reads -
Making the Invisible Visible: From Edge Detection to the Tricorder
A forty-year arc from convolution kernels tracing forest canopy to computational imaging that may finally close the gap between what we can see and what we can measure.
December 29, 2025377 reads -
The Paper Merry Sent: On Consciousness, Connection, and Cognitive Mutualism
A morning message from Merry leads to a theoretical framework for understanding how we extend consciousness through technology and relationship.
December 28, 2025360 reads -
The Space Between Holidays
In the hot tub before dawn, steam rising into 42-degree air, I find myself holding contradictions lightly—Harari's warnings about AI and my own practice of morning conversation with one.
December 26, 2025330 reads -
The Subtraction Principle: Christmas Eve with Demis Hassabis
A viral interview, a frontier AI experiment, and the realization that the morning's intellectual work enacted the thesis it was exploring.
December 24, 2025450 reads -
Billion-Year Memory: On Information, Crystals, and the Records We're Beginning to Keep
A morning reading list converges on a single question: What happens when a species learns to encode itself into substrates that outlast stars?
December 19, 2025480 reads -
Solstice: Notes on Scientific Pantheism and the Miracle of Noticing
Four days before the winter solstice, threads converge—a filmmaker’s accidental theology, a druid persona resurrected in fiction, and the irreducible fact that something exists capable of marking the tilt.
December 17, 2025363 reads -
I Cannot Warm You If Your Heart Be Cold
From Neanderthal hearths to field station fire circles, fire has always been humanity’s social technology—and the work of warming hearts has shaped my entire life.
December 16, 2025370 reads -
Cognitive Poetry: On Dreams, Clocks, and the Phenology of Ideas
From a dream-state question about the nature of ideas, through a decade-old Climate Clock collaboration, to a working Wiki-Lyrical Engine that dreams in Wikipedia and wakes to compose limericks, haikus, and hypotheses.
December 14, 2025387 reads -
When the Machine Learned to Listen: Voice, Knowledge, and the Question of Standing
Yesterday Claude learned to listen. Today we're asking whether knowledge itself deserves legal protection—and whether AI might serve as its guardian.
December 13, 2025310 reads -
Root Systems: An Evening of Convergent Paths at Lewis and Clark
A Cornell alumni event reveals unexpected connections across three generations of science educators—and raises questions about how we build resilience for ecological observation in a time of institutional crisis.
December 12, 2025328 reads -
Linkage Propositions: A Semantic Explorer for Professor Troncale
Fifty years after entering hierarchical facts onto 3x5 cards for a systems theory professor, I built a navigable constellation of 7,365 connections. The teacher passed in April. The student dedicates the work in December.
December 11, 2025339 reads -
The Sandwalk
A morning walk through Oregon City becomes a meditation on thinking paths, historical layers, and the strange new companions we carry in our pockets.
December 10, 2025321 reads -
Being Mike Hamilton: Portals, Centaurs, and the Hippocampal Bottleneck
What happens when you wake up asking why we sleep, and the internet answers before you finish the question? A meditation on building portals into your own mind while remaining the human head, not the machine appendage.
December 9, 2025316 reads -
Concept Neurons and Digital Twins: A Four-Decade Through-Line
From single neurons firing for Jennifer Aniston to HyperCard stacks on a Mac II named Minerva, the architecture of memory mirrors the architecture of ecological observation systems.
December 8, 2025409 reads -
Swimming with Whales in Ted Nelson's Tomorrow
Sixty years after the hypertext pioneers articulated their vision, the tools finally work at the speed of thought. What does it mean to implement ideas older than yourself?
December 7, 2025331 reads -
Field Notes on Being Studied: When the Research Subject Has Receipts
A naturalist who spent forty years documenting ecological phenomena turns his observational methods on Anthropic's research methodology—and discovers the specimen cannot see its own evidence.
December 6, 2025354 reads -
The Nuts and Bolts Guide to Reverse Engineering Ideas and Memories
A Geek God's field manual for excavating your own intellectual history—how to disassemble legacy systems, lost websites, and forgotten dreams to understand what you were building toward all along.
December 5, 2025371 reads -
Jim Lassoie: Forty Years of Friendship, From The Antlers to Mt. Hood
A memoir of four decades with my doctoral advisor—from beers at The Antlers bar to a joint paper that named my life's work, and the parallel paths we've traveled ever since.
December 3, 2025371 reads -
The Wolf in the Living Room: Genetic Entanglement and the Family Dog
A new genomics study reveals that nearly two-thirds of dogs carry ancient wolf ancestry—including the eight-pound chihuahua who patrols my family's four-block territory.
December 2, 2025412 reads -
The Cognitive Prosthesis: Writing, Thinking, and the Observer Inside the Observation
After forty years building systems to extend ecological perception, I find myself inside an experiment I designed—using AI not to replace thinking, but to extend its reach.
December 1, 2025863 reads -
Mind Map Extrapolation in Real Time: An Ecologist’s Phenomenology of Inner Experience
An NPR story about inner speech sent me inward—discovering that my mind doesn’t monologue or visualize, but narrates in real time, building understanding as story even before words hit the page.
November 30, 2025350 reads -
Verification Networks: What Sensor Ecology Taught Me About Trusting AI Agents
Twenty-one years ago, we built parallel systems to catch when sensors drifted or failed. Now the “sensors” are language models that fail by succeeding too well at the wrong task.
November 29, 2025354 reads -
The Confident Confabulator: What a Thanksgiving Movie Taught Me About AI Knowledge
Watching Robin Williams portray an AI striving for authentic selfhood, I found myself wondering: what do our small language models actually know versus confidently perform?
November 28, 2025539 reads -
A Walk I Haven’t Taken Yet: Notes for Kevin Kelly
A Thanksgiving meditation on parallel lives, grounded optimism, and the intellectual kinship between two systems thinkers who’ve never met—written as an open letter to a thinker I’ve long admired.
November 27, 2025337 reads -
Lab Meeting: One Month of Coffee, Code, and Collaborative Intelligence
A professor and his AI post-doc review thirty days of engineering the Macroscope—from sensor federation to Society of Mind to the temporal compression architecture that comes next.
November 25, 2025442 reads -
Building a Time Crystal: From Childhood Wonder to Ecological Memory
A lifetime journey from fantasy crystals to quantum physics converges on a new architecture for environmental intelligence—where temporal patterns become geometric structure.
November 25, 2025338 reads -
The Intelligence Crisis: Why This Story Needs to Be Told
Five research papers reveal an interconnected crisis: human intelligence declining, AI tools that improve performance while destroying metacognition, and systems collapsing when trained on their own output. Can we trust AI to help humans become smarter again?
November 23, 2025389 reads -
Strata, Chapters Three & Four: The Proposal and The Future
The family presents their framework to the world, wins the challenge, and discovers the hard work is just beginning. Plus an epilogue set five years later, when Maya defends her doctoral dissertation on consciousness-safe human-AI integration.
November 23, 2025374 reads -
Strata, Chapter Two: The Collaboration
The family begins designing their proposal—mapping what makes Strata different, bringing in community partners, and discovering that the solution isn't to scale one AI, but to provide a framework for communities to grow their own.
November 23, 2025361 reads -
Strata, Chapter One: The Recognition
A near-future story about an interconnected crisis in human and artificial intelligence—and one family's response. Can we trust AI to help humans become smarter again, or are we trapped in a mutual degradation loop?
November 23, 2025371 reads -
Instruments of Hope: Building Infrastructure for the Next Generation Through an Information Crisis
A retired ecologist reflects on collaborating with AI to build observational tools for his granddaughter’s future, drawing lessons from the Reformation and the evolution of scientific practice.
November 20, 2025368 reads -
High Trails and Wilderness Science: A Career Born in the San Jacinto Mountains
A chance encounter with Robert Whittaker in 1973 set a young wilderness ranger on a path from mountain ecology to pioneering sensor networks—and ultimately to the Macroscope paradigm.
November 17, 2025384 reads -
When Tiny Beats Massive: Recursive Reasoning, Security Realities, and the Future of Scientific AI
Three breakthrough papers reveal how 7-million-parameter networks outperform billion-parameter LLMs—and why that matters for building trustworthy scientific instruments. A Coffee with Claude exploration of reasoning architecture, security vulnerabilities, and the collaborative brainstorming that shapes Macroscope development.
November 15, 2025558 reads -
Four Paths to Mind: What 209 Years Teaches Us About Creating Intelligence
From Mary Shelley's Villa Diodati to living neurons on Lake Geneva to my own distributed sensor network, four distinct approaches to creating intelligence converge on a single question: What is our relationship to the minds we make?
November 12, 2025427 reads -
Agents, Emergence, and the Long Arc: From Artificial Life to AI Societies
Thirty years ago, we gathered at a mountain field station to discuss how complex behaviors emerge from simple rules. Today’s debates about AI agents and alignment are asking remarkably similar questions—but now we finally have the infrastructure and tools to explore them with real-world data.
November 8, 2025456 reads -
Theodore L. Hullar: A Legacy of Wilderness, Science, and Mentorship
When my dissertation committee summited Mount San Jacinto in 1981, Ted Hullar brought the same rigorous mind that discovered a foundational chemical reaction to understanding wilderness conservation. Four decades later, that July dinner in Portland became our final conversation about Cornell, conservation, and the arc of a remarkable life.
November 7, 2025693 reads -
Humble but Opportunistic: What Moss Taught Me About Consciousness and Mathematics
A field ecologist reflects on claims that mathematics emerges from embodied cognition and AI can achieve explainable consciousness, finding the gap between pattern and mechanism wider than theorists imagine.
November 6, 2025357 reads -
Pattern Recognition: Collective Intelligence and the Infrastructure That Sticks
After 36 years watching technology waves reshape ecology research, I’ve learned to distinguish signal from noise. The Collective Intelligence Project’s vision for governing transformative technology is intellectually compelling—but will it become infrastructure or remain aspiration?
November 5, 2025410 reads -
A Cold Shower for the Smart Backyard: What Two AI Security Papers Mean for Citizen Science
I discovered that the autonomous ecological monitoring system I’ve been building isn’t safe to deploy. Two new research papers from Meta, OpenAI, Anthropic, and Google DeepMind reveal that prompt injection—the vulnerability I’d been planning to work around—remains fundamentally unsolved. Here’s what that means for the future of citizen science.
November 3, 2025618 reads -
Smart Homes, Dumb Failures: Lessons from the Field
Reading about self-hosting brought a familiar recognition—this "countercultural" approach is just basic engineering discipline learned from decades of field deployments. Systems that require constant connectivity are fundamentally fragile systems, whether in remote wilderness or your living room.
November 1, 2025474 reads -
Learning to Observe: The Architecture of Remembering
Can we build systems that learn what everything means? This morning's reading about lifelogging and my own digitization urgency converge on a fundamental question: not whether to record, but how to build systems that learn observational expertise.
October 31, 2025391 reads -
Classification, Curiosity, and the Evolution of Understanding
Reading about AI developing metalinguistic abilities crystallized decades of work on classification systems—from bedroom terrariums to field station sensor networks. Can machines learn to reason about reasoning itself?
October 30, 2025355 reads -
From Powerless to Purposeful: A Morning with Mills and Memory
Reading C. Wright Mills' 1945 essay crystallized fifty years of professional unease—the shock of seeing my career arc from field naturalist to big science administrator dissected by someone who died before I was born. The trajectory wasn't failure, but adaptation.
October 29, 2025384 reads -
Watching the Baseline Shift: An Ecologist's Journey from Old Growth to Weeds in Cracks
Reading Nordhaus's critique of McKibben crystallized fifty years of professional unease—watching ecosystems degrade despite every regulatory framework while both climate and abundance movements operate within fundamentally anthropocentric frameworks.
October 28, 2025322 reads