Why landfills are the key to solving AI’s enormous energy needs | Opinion

The world’s leading business leaders are struggling to find the right metaphor to describe the impact of artificial intelligence (AI) on humanity. Some compare it to the printing press, others to the beginning of a new industrial revolution. Sundar Pichai, CEO of Google, even said the AI ​​revolution is comparable to the moment when humanity began to use fire.

Superlatives aside, it is clear that we are in our own Noachian moment – ​​human history will soon be divided into the antediluvian world that existed before artificial intelligence and the world that will take its place.

But there is a major structural obstacle to fully realizing AI’s potential: enormous energy requirements. A recent paper published by the World Economic Forum estimates that the computing power required to sustain AI’s growth rate will double every 100 days. Wells Fargo estimates that by 2030, the additional energy demand of U.S. data centers will be nearly seven times greater than New York City’s current annual electricity consumption.

So where will all this energy come from? And with growing concerns about climate change, how can we ensure that the race for new energy sources does not hasten the end of our planet?

Goldman Sachs predicted that nearly half of new energy supply will need to come from renewable sources, yet wind and solar technologies are nowhere near enough to meet AI’s enormous needs. But there is one source of renewable energy that remains largely untapped, despite attracting increasing interest from Wall Street to Silicon Valley: waste streams.

You might call it garbage. The British know it as scrap. Construction companies might call it rubbish. But we should all probably call it something else – unused energy.

“Humanity will always produce waste, whether through household waste or through large amounts of waste from industry,” says Cornelius Shields, founder and CEO of Kore Infrastructure, one of America’s leading renewable energy innovators that uses non-combustion conversion technologies to provide clean energy to public utilities like SoCalGas, Southern California’s primary natural gas provider. “There will always be huge amounts of waste on Earth and compared to wind and solar energy, waste streams are relatively easy to capture. By using technologies that sustainably convert biogenic waste into clean energy, namely renewable hydrogen and biocarbon, we can produce energy at scale while reducing harmful emissions that contribute to climate change,” he says.

If using ongoing waste streams for clean energy sounds like a good idea to you, imagine what that could mean for landfills—a constant concern for regulators. Depending on how you define “landfill,” there are as many as 2,600 such sites in the U.S. alone. According to the EPA, “US landfills released an estimated 119.8 million metric tons of carbon dioxide equivalent into the atmosphere in the form of methane in 2022.” Kore Infrastructure estimates that the potential energy diverted from these landfills could meet up to half of the currently estimated electricity needs of AI data centers.

And that only in the USA

“Many landfills are at capacity and the waste is being transported further and further away, which in turn leaves its own carbon footprint,” Shields said. “But imagine if waste streams that would normally end up in landfills were not a problem, but actually became an energy source to power AI data centers.”

Using green waste-to-energy technologies to generate energy for AI around the world could transform huge centers of waste production—places like Mumbai, Manila, and Mexico City—into the energy hubs of the future. The net result would not only power energy-hungry AI data centers around the world, but also curb dangerous carbon emissions at the same time.

Ironically, global demand for the waste streams required for the AI ​​revolution may inadvertently trigger the market economy needed to combat global warming on a large scale.

Technology observer Gene Munster, Wall Street’s leading analyst on all things AI, agrees. He predicted, “There’s no question that current energy supplies are nowhere near enough to power AI beyond 2027, and unless there are major breakthroughs in new energy sources, there will definitely be some throttling. But the economics of AI are too compelling to let that happen, which is why I think Silicon Valley’s need for (alternative) energy sources could well open a back door to address climate change in a way that makes economic sense. That would be a huge win for everyone. And for the planet.”

Shields agreed and said Newsweek that “capturing society’s waste streams and using them to meet the enormous energy needs of the future in a way that also takes climate change into account is not just wishful thinking. It is a necessity.”

“It may be a somewhat ironic twist when we realize that AI, one of humanity’s greatest technological achievements, could also be the unintended force that enables us to save our planet,” he added.

Arick Wierson is a six-time Emmy award-winning television producer and was Senior Media and political advisor to former New York Mayor Michael Bloomberg. He advisesCorporate clients with communications strategies in the USA, Africa and Latin America.

The views expressed in this article are those of the author.