research, running and recompense
I was listening to a podcast this morning recommending a book representing one of the best deals on Audible (if you are a subscriber). The book is the Power Broker. It is a bazillion pages but so inexpensive as an audio listen that I ran an entire summer in NYC listening to each chapter and then a recap over at 99% Invisible.
The interesting thread to the rest of what I am sharing is how building density and access for transportation in NYC had the opposite impact than what was anticipated. Not really a big leap to AI infrastructure expanding 10x and may not take us where we think we are headed.
The Power Broker is first and foremost a brilliant multidimensional portrait of a man—an extraordinary man who, denied power within the normal framework of the democratic process, stepped outside that framework to grasp power sufficient to shape a great city and to hold sway over the very texture of millions of lives. We see how Moses began: the handsome, intellectual young heir to the world of Our Crowd, an idealist. How, rebuffed by the entrenched political establishment, he fought for the power to accomplish his ideals. How he first created a miraculous flowering of parks and parkways, playlands and beaches—and then ultimately brought down on the city the smog-choked aridity of our urban landscape, the endless miles of (never sufficient) highway, the hopeless sprawl of Long Island, the massive failures of public housing, and countless other barriers to humane living. How, inevitably, the accumulation of power became an end in itself.—Amazon, author Robert A. Caro
I am familiar with Jevons paradox — usually applied to energy efficiency models.
I enjoyed the segue over at Linear Digressions with Katie Malone.
Jevons Paradox is the frame you need for AI spending right now. When the per-unit cost of a resource drops dramatically, usage tends to rise even faster, so total spending goes up. This pattern, first observed with coal in the 1860s, is playing out almost textbook-style with LLM inference: per token costs down a thousandfold, total AI spend up roughly tenfold since 2022.—
What does this have to do with open-source geospatial? There is so much to learn about AI infrastructure. There are cost paradigms and adoption rates to consider. Here is a photo dump of some of the graphics for stories I am working on to share from the stage.
talks:
Drawing the Lines: Visualizing the role of climate change and urbanization shaping human-bear interactions in Japan
Not on Your Google Maps: The Map Is Not the Territory
workshop:
Boundless Demand, Bounded Earth: A 3D Look at Cities and Resources
Invited keynote speaker
Cohab.space — in conversation and workshop facilitator (storytelling + visual perspective)
click to join the conversation…
