Is AI Burning the World?

Midjourney; Curt Doty

Green AI: Why New Mexico Might Just Save the Future

Let’s set the record straight: AI isn’t the planet-scorching Skynet some headlines make it out to be. Yes, AI needs energy. A lot of it. But so does crypto, Netflix, and your always-on Ring doorbell.

The question isn’t whether AI uses energy. The real question is how we power it. And here’s where New Mexico enters the chat.

Is AI Really an Energy Monster?

Sure, training large models like GPT eats up kilowatts like TikTok chews through your screen time. But the alarmism misses the mark. Data centers already exist. AI workloads are increasingly replacing—not adding to—legacy compute. And unlike fossil fuel exploration, AI has an actual upside: it might just help us optimize the entire energy grid.

Imagine AI not as the villain, but as the thermostat—regulating energy demand, predicting grid failures, and accelerating the transition to renewables. It’s already happening. So maybe let’s cool it with the doomscrolling.

AI Energy Use Case Table

Sources: Kanoppi CreaUnited EpochAI Brookings

China’s Power Play: Green Energy + AI

While we’re still writing policy memos and throwing venture capital at vaporware, China is aligning its green energy investment with AI deployment like it’s a national strategy. Solar-powered data centers? Check. Subsidized chips? Yup. Rural provinces turning into AI hubs? Already happening.

The West is still crowdsourcing solutions. China is scaling them. That’s not just an edge—it’s a full-court press. And it should terrify us… unless we act smart.

New Mexico: The Untapped Hotbed of Green AI

Now to the good part: New Mexico. It’s not just the land of chile and sunsets anymore—it’s poised to be the Silicon Mesa of sustainable AI.

You’ve got the usual suspects: abundant sun, strong wind corridors, and a growing commitment to clean energy. But here’s the sleeper hit: geothermal.

New Mexico has one operational geothermal plant. Just one. But the state sits on a treasure map of untapped geothermal hotspots—an underground energy goldmine, just waiting for the right investment. Unlike solar and wind, geothermal is baseload power—24/7, consistent, infinite.

In a world chasing battery breakthroughs and weather-dependent grids, geothermal could be New Mexico’s ace.

The Quantum Connection

But wait, there’s more. Quantum computing—the next frontier—is quietly gaining traction in New Mexico. While AI consumes energy, quantum could one day rewrite the equation with exponentially faster, less power-hungry computation.

The latest quantum deal in New Mexico is a major partnership and investment announced in early September 2025. The state of New Mexico, in collaboration with the Defense Advanced Research Projects Agency (DARPA), launched the Quantum Frontier Project—a $120 million, four-year effort to accelerate quantum computing research, testing, and independent verification. This partnership is part of the national Quantum Benchmarking Initiative, leveraging expertise from the state’s universities, national laboratories, and private sector. (Source: Reuters)

On top of this, New Mexico committed a total of $315 million to develop quantum infrastructure, invest in private companies, manufacturing facilities, and create a quantum network across the state. Of this, $185 million will be allocated from the state’s sovereign wealth fund to enable private venture capital investments in quantum businesses operating in New Mexico. These moves are designed to position New Mexico as a national leader in the rapidly growing field of quantum technology. (Source: QuantumInsider)

Pair geothermal + quantum + AI, and you don’t just get a green energy solution—you get a blueprint for the future.

Data Centers

OpenAI is building a massive new data center in southern New Mexico (Santa Teresa, Doña Ana County) as part of its $500 billion Stargate initiative in partnership with Oracle and SoftBank. This is one of five new hyperscale data centers being developed in the United States, each designed to dramatically expand OpenAI’s compute infrastructure for training and operating advanced AI models. (Source: Reuters)

The Southern New Mexico facility, called Project Jupiter, is a $165 billion campus—the largest single AI data center investment announced for the border region to date. Its selection followed a nationwide search, with the region chosen for its proximity to power resources and fiber networks and the ability to support high-density, liquid-cooled compute racks. (Source: BizJournals)

Reframing the AI Conversation

Right now, the public narrative around AI is a mess: it’s either going to steal your job or your soul. But what if we framed it differently?

What if AI is our tool to fight climate change—not another contributor to it? What if New Mexico became the unlikely proving ground for sustainable AI? Less Blade Runner, more Solarpunk.

Questions Worth Asking

• What are the real energy costs of AI—and how do they compare to, say, global aviation?

• What policies will turn New Mexico into the Iceland of the Southwest?

• Can AI + geothermal energy turn climate guilt into climate action?

(Short answer: Yes. Long answer: If we let it.)

Bottom Line

We’re at a crossroads. We can either keep burning resources to fuel our digital lives—or we can drill into a future that’s been under our feet all along.

New Mexico’s not just ready. It’s uniquely positioned. With its natural assets, research institutions, and quiet momentum in geothermal and quantum—it might just lead the green AI revolution.

The future of AI doesn’t have to be carbon-stained. It can be clean, conscious, and powered by the sun, the wind, and the core of the Earth itself.

Table Notes

Equivalent household hours: The average U.S. home uses about 1.2 kWh per hour, or 10,500 kWh per year.

Water use: Cooling AI computations typically requires about 2 liters of water per kWh—used for data center cooling.

Metaphors:

Writing a blog or doing research with AI consumes enough water to fill a large glass—about what a person drinks per sitting.

Creating an image with AI consumes enough water to fill two bathtubs.

Generating a short video (5 seconds) with AI uses enough electricity to run a household for about a day, and enough water to fill a small swimming pool.

Sources

Analysis by Kanoppi and ACM FAccT 2024 show a single ChatGPT query uses about 0.0029 kWh, while a Google search is about 0.0003 kWh.

The water footprint of AI, estimated at 2 liters per kWh, is drawn from CREA United and World Economic Forum data.

Video generation by advanced models can use more than 3.4 million joules per task, roughly 0.94 kWh per 5-second generation.

Household energy consumption metrics and equivalence calculations based on U.S. Department of Energy stats and EcoFlow metrics.

About the Author

Curt Doty, founder of CurtDoty.co, is an award winning creative director whose legacy lies in branding, product development, social strategy, integrated marketing, and User Experience Design. His work of entertainment branding includes Electronic Arts, EA Sports, ProSieben, SAT.1, WBTV Latin America, Discovery Health, ABC, CBS, A&E, StarTV, Fox, Kabel 1, and TV Guide Channel. His work in movie marketing spans the major studios: Universal Pictures, Fox Searchlight, 20th Century Fox, Lionsgate, Miramax and Disney. He is now helping independent filmmakers market their movies for festivals and distribution.

He currently serves on the board of the Godfrey Reggio Foundation and is the AI Writer for Parlay Me.

To learn more about Curt’s pedigree of innovation, check this out.