
AI’s power problem
AI’s Insatiable Appetite: The Looming Energy Crisis and the Path to a Sustainable Digital Future
(This article was generated with AI and it’s based on a AI-generated transcription of a real talk on stage. While we strive for accuracy, we encourage readers to verify important information.)
Mr. António Coutinho, CEO of EDP Innovation, addressed “AI’s power problem” at Web Summit Lisbon 2025, highlighting its significant, often-overlooked energy dependency. He emphasized that every AI operation relies on physical infrastructure, asserting “no energy means no intelligence.” Since 2014, AI compute has surged 350,000 times, while global energy generation assets increased by only 1.56 times, creating a critical imbalance.
This exponential digital growth clashes with the incremental pace of physical energy infrastructure. Tangible bottlenecks are evident, with grid capacities maxed out in regions like Ireland, Singapore, and North Virginia, impacting thousands of businesses. Data centers in Ireland now consume 20% of the country’s energy, delaying new connections until 2028. Power prices are escalating, making energy a critical constraint.
Each hyperscale data center requires 300-500 megawatts, doubling with cooling, equivalent to a medium-sized power plant. As energy becomes a strategic resource, geopolitics shifts towards control over grids, storage, and generation assets. Only 26% of the world is energy independent, leaving 74% reliant on fossil fuel nations.
Europe, importing 60% of its energy, faces a substantial economic disadvantage compared to the US, which gained a $6 trillion digital investment advantage due to its energy independence. The path to energy sovereignty and autonomy lies in electrification, especially as 92% of the global population resides in areas rich in renewable resources.
Electrification is both cleaner and significantly more efficient. Electric vehicles consume four times less energy, and heat pumps three to four times less than traditional counterparts. Global investment in clean technology reached $2.1 trillion last year, with China contributing 40% and rapidly deploying solar capacity.
Energy leadership is crucial for winning the AI race. Renewables offer compelling economic advantages: costs for solar, wind, and storage have plummeted by 65% in recent years, making them cheaper and faster to deploy than fossil fuels or nuclear. A 1 GW solar plant can save $250 million in gas imports, with a payback period of just 3-4 years.
Despite technological readiness, systemic hurdles persist, including lengthy permitting processes (7-10 years for wind farms). Modernizing a 200-year-old energy infrastructure in 25 years demands significant innovation, encompassing standardization, prefabrication, and automation. EDP Innovation, through programs like Free Electrons, supports startups building this essential infrastructure, recognizing the “next unicorn” will emerge from this intersection.
Mr. Coutinho concluded that energy is the new bandwidth, forming the fundamental foundation of all digital ambition. AI’s future is intrinsically linked to the electrons we can generate and deliver. The power won’t just be coded; it will be powered by a transformed, sustainable energy infrastructure.

