If artificial intelligence is going to sap resources and strain the grid, it better produce results that justify those consequences. So far, the technology seems to be heavily weighted by the bad.
To be sure, there are good developments and outcomes in the field, and they might be underplayed. Take its ability to help address food insecurity, improve extreme weather preparations, and get the most out of solar panels.
However, extreme energy consumption, water use, and pollution have to be considered in the equation. An AI data center can burn through as much power as tens or even hundreds of thousands of homes with "volatile" consumption patterns that threaten grid "harmonics" and could lead to appliance failures, electrical fires, and blackouts and brownouts in residential homes, as Bloomberg detailed.
Data centers also rely on vast quantities of water and produce air and noise pollution that harm communities and the environment.
What is especially concerning is that a large data center can use as much water in one day as 4,200 people, according to the University of Illinois Urbana-Champaign. Water is a nonrenewable resource, with only 0.5% of freshwater on the planet available to humans, and 1 in 10 people worldwide already live in water-stressed countries.
Big Tech is supposedly turning to clean energy, such as solar and wind power, to run its data centers, which don't exclusively power AI, but the energy industry's slow transition from coal and other dirty fuels means they are still a significant source of pollution.
Watch now: Would you wear a shirt made out of oysters?
U.S. data centers contributed 105 million metric tons (over 115 million tons) of carbon dioxide to the atmosphere in the year ending in August, MIT Technology Review noted, and the sector is more carbon-intense than the average. And then there's the Guardian analysis that showed data center pollution is over seven times greater than reported.
The problem mushrooms when you consider what all this is turning out. Recently, Meta removed AI profiles from its platforms after receiving backlash. The characters were created in 2023 but garnered attention when an executive detailed company plans to have AI populate Instagram and Facebook like real people.
The experiment continued the "digital blackface" spectacle that has infiltrated the internet for over 10 years, as Karen Attiah of The Washington Post wrote. Attiah shared offensive and concerning responses from a chat with one of the profiles, called "Liv," and charged Meta with trying to boost engagement via the manipulation of marginalized groups.
Do you worry about companies having too much of your personal data? Click your choice to see results and speak your mind. |
"Even when controlled by engineers, these AI profiles struggled to handle nuanced questions from skeptical users," Techstory added. "Liv's inability to address cultural and racial identity issues raised serious ethical concerns and highlighted the risks of AI reinforcing biases."
The practical failure of the bots was one thing — Liv told different users different tales about its origin story and made up information about its creators — but the ethical and cultural violations underscored the issue.
Facebook has long toyed with human emotions and even "actively suppressed" Black and queer dialogue, as Attiah said, but now, it's venturing into territory that even the most ardent AI supporters can't defend.
"The bottom line: No one wants this. No one needs this," Attiah wrote. "Meta, with its incredible power over what billions of people around the world see, is willing to do nearly anything to keep us addicted to its platforms — even if that means flooding the zone with digital slop that doesn't work very well. And worse, it's digital slop that can cause serious harm by reinforcing cultural biases and stereotypes."
Join our free newsletter for good news and useful tips, and don't miss this cool list of easy ways to help yourself while helping the planet.