When French President Emmanuel Macron stood on stage at the Paris AI Action Summit and dropped the mic with the phrase “Plug, baby, plug”, it wasn’t just a catchy slogan — it was a rallying cry. A statement that France, and potentially Europe, intends to build the future of Artificial Intelligence on a foundation that’s not just powerful, but sustainable.
At a time when other global powers lean harder into fossil fuels to power their AI ambitions, Macron offered an alternative vision: one that sees sustainable energy and advanced computing not as competing priorities, but as mutually reinforcing imperatives.
France's Macron makes closing remarks at first day of AI summit in France (February 10th 2025)⚡ The power behind the intelligence
Let’s be clear — AI is energy-hungry. Every model we train, every prompt we generate, and every data center that supports real-time inference draws on an immense pool of power. As AI adoption scales globally, so does its carbon footprint — unless we make conscious choices about where and how we source the electricity that feeds it.
Here’s the choice laid out before us:
- Option A: Build AI on a bed of coal, oil, and political convenience. It’s fast, cheap (for now), and dirty. This is the path being doubled down on in places where climate commitments have taken a back seat.
- Option B: Build AI on clean grids — solar, wind, hydro, and yes, nuclear. Slightly slower at first, potentially more expensive, but infinitely more resilient and ethically sound in the long run.
Macron’s France is opting for Option B — not just for environmental reasons, but for economic, geopolitical, and technological sovereignty as well.
🇫🇷 France’s “Notre-Dame strategy” for AI
The so-called “Notre-Dame strategy” references the rapid and coordinated rebuilding of the fire-ravaged cathedral. Macron is calling for that same urgency in constructing a European AI infrastructure: modular, sovereign, and green.
A few highlights from the summit:
- €109 billion in investment to accelerate AI infrastructure across Europe.
- New clean-powered data centers backed by energy from France’s extensive nuclear grid.
- A public-private push to keep Europe in the AI race without outsourcing its digital soul to the U.S. or China.
This isn’t just politics. It’s design thinking at the national level: How do we shape an AI future that doesn’t burn out the planet while training the next model?
🌍 AI shouldn’t cost the Earth
Contrast that with the vision being put forward in other parts of the world — particularly in the U.S. under certain leadership visions where climate policy is seen as a burden, and fossil fuel resurgence is framed as a patriotic act. In those circles, AI innovation becomes yet another excuse to postpone a transition we all know is overdue.
But AI shouldn’t cost the Earth — literally or figuratively.
As we ask models to make better decisions, summarize vast documents, or generate synthetic content, the question behind the question becomes:
How is this intelligence powered?
And maybe more importantly:
Is the intelligence we’re building also sustainable in a moral, ecological, and democratic sense?
🔌 A call to build differently
“Plug, baby, plug” isn’t just about electrons. It’s about values.
It’s a call to plug into smarter infrastructure, plug into ethical design, and plug into collaboration over extraction.
It’s about recognizing that the race to build better AI is inseparable from the race to decarbonize our economies. And if we get this right — if we invest in clean compute, open standards, and public-good datasets — then Europe (and other like-minded regions) can lead not just in AI capabilities, but in defining the moral framework around it.
🚀 Final thoughts
The future of AI won’t be decided only by parameter counts, token throughput, or funding rounds. It will also be shaped by the principles we encode into the very power sources that sustain it.
So yes — plug, baby, plug. But make sure what you’re plugging into is clean, just, and visionary.
"Emmanuel Macron's closing remarks at first day of AI summit in France (February 10th 2025)" by AI Summit.