In the race to build (or stop building) data centers across the country, it's important to remember that no government is spending more on artificial intelligence than the United States. In fact, according to the Brookings Institute, the number of AI contracts within the federal government has increased from 472 in 2022 to more than 1,700 in 2026.
Most of the spending ($90 billion) is by the Department of Defense, however other departments (Commerce, Health and Human Services, NASA) are spending tens of millions. So regardless of how one may feel about AI, we are all tied to its use.
The site locations and economic benefits of data centers can be debated, but there is no getting around the fact that they are the physical embodiment of what AI does. The more computing power a model needs, the more square footage it will need to build.
What worries me most is not the size of the buildings or the energy consumption. Isn't it knowing whether we're looking at a disruptive technology that will ultimately help all of society, or a disruption that exists primarily to make the richest people on the planet even more money? I struggle with that question because our history is peppered with examples of the latter.
The world's first billionaire, for example, became a household name in part because demand for electric cars skyrocketed shortly after electric cars all but left the market. Confused? Yes, me too. General Motors spent years developing its EV1 and then abruptly discontinued the model, citing production costs. Was it really that or the oil industry, through campaign donations and lobbying, undermining the success of the automobile to protect its own interests? Or were we, the public, the ones who didn't want to change? Who knows for sure?
One thing I do know is that Larry Burns, the former vice president of research and development at General Motors who killed the automaker's electric car, said that if he could go back in time, they wouldn't have stopped. invest in emerging technology.
I remember those 32 solar panels President Carter had moved into the White House back in 1979. They didn't make it past the next administration. Today China has more than 80% of the solar market — of each stage of production. The United States not only sacrificed its opportunity to corner a future clean energy market when we were its pioneers; We also gave up all manufacturing and supply chain jobs that would have come with that domain.
We didn't lose that race. We lost it.
Let's go back further, to 1937 and the Marijuana Tax Law. It was not based on science. It was built on fear, newspaper sensationalism and, some historians maintain, the interests of industries that did not want hemp fiber to compete with their own products. Science separated THC from fiber, but the story elected officials told the public did not. If hemp, the building material, and marijuana, the drug, had not suffered the same fate, perhaps it would be the United States and not Europe that would be the leaders in industrial hemp production. Unfortunately, due to the stigma associated with the plant, industrial hemp production was illegal for 80 years, until the 2018 Farm Bill.
Today, as American hemp producers struggled through decades of federal turmoil only to return to the field, Belgium's IsoHemp runs an automated factory that produces millions of blocks of hempcrete a year, and Europe's industrial hemp construction market is projected to grow nearly sixfold by 2033.
“There are so many historical uses for the fiber, from ropes to textiles, that it's really hard to understand how damaging the ban was to the science of the industry,” said Mattie Mead, who founded Hempitecture in 2013. Based in Idaho, Mead's company produces insulation and structural building material that incorporates hemp fibers and a lime-based binder. Using hemp in building construction provides an alternative to fiberglass, helps reduce carbon emissions, and naturally keeps buildings cooler in summer and warmer in winter, reducing energy costs.
“Things are a lot better now than when we started, but there's still a lot of stigma and misinformation,” Mead said. “It's as if the public has to unlearn the myths and falsehoods first before they can see how the technology helps.”
The same can be said for data centers and AI.
Much of the technology has been involved in apocalyptic scenarios and Terminator movies. Sometimes it is difficult to separate the impossible from the improbable. Even harder if you can't tell if the technology is intended for the good of everyone or simply intended to be good for someone's bottom line. Perhaps, unlike the “marijuana” prohibition era, the dire warnings are valid. But now, as then, we need to separate fact from myth before setting policy.
YouTube: @LZGrandersonShow
