AI's voracious power consumption fuels electricity demands

Recently, Bill Gates discussed a range of subjects with a small group of reporters, including causes, possible remedies, and climate change. It's interesting that Gates brought up the energy usage of artificial intelligence.

According to Gates, AI data centers generate only a minimal amount of extra stress for the grid. Furthermore, he projected that efficiency improvements from AI-derived insights will more than offset that higher demand.

That's a bullish opinion from a billionaire with personal stakes in the outcome. As the former CEO of Microsoft and a significant shareholder in the firm at the forefront of the AI revolution, Gates is a huge climate investor.

Even though it's too soon to make a firm judgment on the matter, a few facts are evident already: With no end in sight, artificial intelligence (AI) is having a significant impact on global energy demand and frequently contributing to an increase in emissions that overheat the earth.

AI data centers need a lot of electricity. Compared to a typical microprocessor, the so-called graphic processing units, or GPUs, that are used to train huge language models and answer ChatGPT inquiries consume more energy and produce more heat.

The number of data centers going online virtually every week is driving up estimates of the amount of energy needed to fuel AI growth.

A peer-reviewed study estimates that artificial intelligence (AI) might supply 0.5% of the world's electricity by 2027, or around what Argentina uses per year. Wells Fargo experts predict that by 2030, artificial intelligence will be a factor in a 20% rise in the nation's electricity consumption.