Artificial intelligence, while heralded as the next frontier of innovation, carries an increasingly heavy and often unseen energy burden. The rapid proliferation of AI tools and services is poised to double the electricity demand from data centers in the U.S. by 2028, a staggering increase equivalent to powering an entirely new state.
This critical, yet opaque, aspect of AI's growth was the subject of a recent CNBC 'Power Lunch' interview with Joanna Stern, Personal Technology Columnist at The Wall Street Journal, who shed light on the industry's lack of transparency regarding its escalating energy consumption.
Stern's investigation highlights a fundamental challenge: quantifying the energy cost of individual AI interactions is remarkably difficult. "It's really hard to figure out how much energy each of our AI prompts take," she revealed, pointing to a pervasive lack of transparency from major tech companies developing these advanced systems. While basic text queries like those on ChatGPT are relatively low-impact—Sam Altman noted the average ChatGPT query as "0.34 watt-hours"—more complex AI-generated content demands significantly more power.
The disparity in energy usage between different AI functions is stark. For instance, generating a single six-second AI video can consume "anywhere between 20 watt-hours and 110 watt-hours." To put this into perspective, Stern offered a relatable analogy: creating one or two such videos is akin to grilling a steak on an electric grill.
The true scale of AI's energy footprint remains largely hidden from the end-user. Unlike electric vehicle owners who directly observe the impact of charging on their home electricity bills, the energy consumed by cloud-based AI operations is absorbed by massive data centers. This creates a dangerous disconnect, where the convenience of AI masks its profound environmental and infrastructural implications. The industry itself acknowledges the concern, with Stern noting, "We've seen a lot of freak out in the AI industry about how much water am I using, how much energy am I using, but we don't really quite know."
This lack of quantifiable data and transparent reporting poses a significant challenge for policymakers, investors, and the public. As AI continues its explosive growth, the demand for electricity will only intensify, placing immense pressure on existing power grids and potentially leading to higher energy costs and increased reliance on carbon-intensive sources if not managed proactively.
