I’m a child of the 1940s and 1950s, a simpler time. Back then, movie and TV westerns made it easy to spot the good guys. They wore white hats. The bad guys wore black.
Fast forward to today. What color is AI’s Stetson? Today’s news and social media present a negative narrative. They remind us, day in and day out, that chatboxes lie, hallucinate, and pretend to hold answers when they don’t. They’re laying bare our secure information, robbing our bank accounts, and numbing our kid’s brains, swooshing them down a slippery slope of generational cognitive decline. Data farms are multiplying like rabbits across the country, guzzling energy, consuming water, and mutilating the rural landscape, all the while raising the specter of a financial bubble. Though apparently increasing employment in the short term (someone has to build out those data farms, write that training code), AI is painted as threatening entry-level employment across all professional fields. AI-enabled robots will take over the factory jobs. AI will fall into the hands of terrorists and bad actors who will create novel biological viruses, compromise our critical infrastructure and use other means to wage war and upend life as we know it. Or AI will go rogue and do those same things on its own.
Whoa! “Roy, Gene, we’re done fer.” (Any readers old enough to remember these white-hatted cowboys?)
These days, Artificial Intelligence could use rebranding—could use a white hat.
Life’s haberdashery offers several. Take medical diagnosis. AI knows the medical literature, and stays abreast of the thousands of new publications and studies added every day. It can catch patterns and connections humans might miss. It offers speed and accuracy. For you and me as individuals, such AI-enabled features of healthcare might someday prove priceless.
Meteorology and related geosciences and their applications offer another—this of existential value to life as a whole, both human and otherwise. Earth’s ability to support a broad spectrum of life is vast, but ultimately finite. It’s dependent on human ways of doing business and corresponding linkages to global and local ecosystems and energy, water, and other natural resources. We can gulp and squander resources while causing great environmental harm. Or we can concentrate on renewables, sip and recycle the rest, and go easy on ecosystems and habitat. The interdependencies are complex and as yet their emergent, long-term consequences remain incompletely identified and understood. All this ignorance combines to compromise effective decision-making. None of society’s practices is sustainable. We can only buy or lose time (the thinking behind the Bulletin of the Atomic Scientists’ doomsday clock).
How much time do we have as the human race? Answers are varied and uncertain. But estimates could be sharpened with better characterization, analysis, and understanding of all the bits and pieces, as well as the whole of the “conceptual diagram” of the Earth system reproduced in the previous LOTRW post. And the estimates aren’t immutable. Any problem that crops up can and will stimulate creative ideas for eliminating or sidestepping it. How much time we have depends on how smart we are.
The good news is we don’t have to create new initiatives for learning more about the Earth and making more efficient use of it. Such efforts are already underway
Given the NASA origin of Bretherton’s conceptual Earth system diagram, it shouldn’t be surprising to find agency follow-up. Case in point, NASA’s Earth System Digital Twins (ESDT):
Organized around interconnected, multi-domain, high-scale modeling capabilities, the three major components of an ESDT are a continuously updated Digital Replica of the Earth System of interest, dynamic Forecasting models and Impact Assessment capabilities.
The Digital Replica is fed by continuous and targeted diverse observations, powered by Data Assimilation and Fusion, and provides an accurate representation of the current state of the system.
Forecasting is facilitated by advanced computational capabilities, Machine Learning (ML) and Surrogate Modeling, and provides real-time or near-real-time prediction of future states of the system.
Finally, Impact Assessment uses the Digital Replica and forecasting capabilities with ML, causality, uncertainty quantification and advanced computation and visualization capabilities for running large amounts of simulated predictions quickly and at various spatial and temporal scales, opening the door to investigating “what-if” scenarios. These will allow [sic] to understand better the impact of one system on another one, and better predict Earth Science events, e.g., severe weather events. This third component provides hypothetical future states of the system.
These same three generic elements (Where are we? Whither are we tending? What could/should we do?) feature in two other efforts. The European Union is coordinating its own version under the label Destination Earth: “Destination Earth is a flagship initiative of the European Commission to develop a highly-accurate digital model of the Earth (a digital twin of the Earth) to model, monitor and simulate natural phenomena, hazards and the related human activities. These groundbreaking features assist users in designing accurate and actionable adaptation strategies and mitigation measures.”
Unsurprisingly, Nvidia itself (the subject of the previous LOTRW post) also has a suite of such activities under the label Nvidia Earth-2:
NVIDIA Earth-2 is a comprehensive family of open models, libraries, and frameworks that democratize global access to professional-grade weather and climate AI. By integrating state-of-the-art architectures, the platform provides a fully open software stack that accelerates every stage of forecasting, from data processing to high-resolution visualization. This shift from resource-intensive physics models to an end-to-end AI pipeline lets researchers, startups, and government agencies run, fine-tune, and deploy sophisticated forecasting systems on their own infrastructure with unprecedented speed and cost-efficiency.
Careful reading of these scraps of text and their underpinnings illustrate these programs are open-ended when it comes to practical application of the forecasts. For example, Nvidia accomplishes applications through what it calls a Partner Ecosystem. Nvidia also offers a small number of Earth-2 Demos labeled From Global Insights to Local Action, with this invitation: “Explore interactive workflows to see how Earth-2 models and tools transform planetary data into actionable intelligence for energy, safety, and infrastructure.”
Note that all three of these approaches are multiscale; they address the Earth system as a whole, but also downscale to local forecasts and to local-scale applications.
The latter should be of particular interest. The October 3rd-9th print issue of The Economist contains an op-ed by Alexander Heffner. He argues that “mayors have become democracy’s most reliable defenders” in the face of authoritarian pressure at national levels worldwide. He states the mayors can stand strong because younger, more diverse, more educated voters live there. He cites Harvard’s Edward Glaeser that today’s cities are engines of innovation and that “mayors, not states, run the country’s laboratories of democracy.” Though Heffner focuses on America, he emphasizes that “around the world, mayors have become the most reliable bulwark against authoritarian politics.”
Strong stuff! For meteorologists and scientists in related fields, looking for meaningful work and stable support as time-honored federal-level opportunities shrivel up or disappear completely, opportunities beckon. Plug in to existing programs such as these three. Remember, they want your ideas; they’re struggling to get in the public’s good graces. They want to be crowned with that big white hat.
Or form your own entity in someone’s Partner Ecosystem. Pay special attention to possibilities at the city level (The United States Conference of Mayors, the National Association of Counties, and the Global Parliament of Mayors might be useful entry points.) Again, the AI should be especially interested in a future rich in myriad activity at the city level, offering opportunities for early detection of success and value and generating data that could be mined for even faster learning about effective local-level policies for sustainable development in the future.
Check out the existing opportunities. Create new ones! Close to home, keep in mind that AMS’s extensive local footprint, comprising forecast offices, broadcast meteorologists, K-12 educators and university faculty, and local AMS Chapters offer intriguing potential assets.
And while you’re at it, get yourself measured for your own (white) hat size. The world needs your insights and will appreciate and recognize your help.
