Digital technology is rapidly reshaping the world around you, but it’s also demanding a massive amount of energy. One question that’s growing louder is whether renewable energy can keep up with this digital boom.

As more people rely on AI, cloud storage, and cryptocurrencies, energy consumption is becoming a concern. Coming to you from Vox, this insightful video tackles the growing challenge of how our digital lives are increasing energy consumption at a staggering rate. The video emphasizes that data centers, which power much of our online activity, are energy-intensive. In 2022, the electricity consumption of data centers, AI, and cryptocurrencies was about 2% of the total global demand. That number is expected to double by 2026, making it comparable to adding the energy usage of an entire country like Sweden. This trend has significant implications as the push for green energy clashes with the rising demand from data-heavy technologies.

The video sheds light on the fact that companies like Google and Microsoft have made commitments to use more green energy in their data centers, but meeting those goals is not straightforward. Renewable energy sources like solar and wind are not always available, yet data centers need to operate around the clock. Building a data center can take a year, while developing new renewable energy infrastructure may take several years, creating a mismatch. Additionally, tech giants aren’t entirely transparent about how much of their energy use is devoted to AI workloads. This opacity complicates the picture and makes it harder for researchers and policymakers to gauge the true impact of these technologies.

The video also discusses some specific figures and scenarios that highlight the complexity of this issue. For example, a single large language model, like GPT-3, consumes about as much electricity in training as 130 average U.S. households do in a year. And AI models are growing fast—the energy needed to train these systems is expected to double every nine months. This growth doesn’t even account for the energy used when people interact with AI applications. On an individual level, one GPT interaction consumes a small amount of energy, but millions of interactions add up quickly.

There’s a tension here that companies and policymakers have to navigate. The video raises the point that, while tech companies may promise to go green, data centers are still relying on fossil fuels as backup energy sources. This reliance could lead to unintended consequences, like coal plants staying operational or new natural gas plants being built to support increased energy needs. In short, even if tech companies meet their sustainability goals, the global energy mix may still lean on fossil fuels, undermining broader climate efforts. For a more detailed look into how AI, data centers, and cloud storage are shaping the future of energy consumption, check out the video above for the full rundown.