While California burns, a scientist from the world’s biggest artificial intelligence chipmaker is suggesting that AI could help fight the wildfires — even though the technology has already begun degrading the environment, and is slated to do much more.
In a post on X-formerly-Twitter featuring a video of a humanoid robot walking down a street in China, Nvidia “embodied AI” researcher Jim Fan suggested, after complimenting the bot’s “swag,” that such machines need “to go fire-fighting asap!”
I’ve verified from reliable sources. New visual Turing Test: this is a real robot demo! The controller is a neural net trained in Isaac simulator using reinforcement learning and then sim2real. Reward engineering is all you need.
Walking gait’s got swag but we need these robots… https://t.co/fwAjLPd51g pic.twitter.com/w3CxFKoRNy
— Jim Fan (@DrJimFan) January 9, 2025
Unfortunately, that tone-deaf suggestion isn’t the only Nvidia blunder that occurred during this week’s fires in Los Angeles.
As the gaming site Aftermath notes, this year’s Consumer Electronics Show took place in Las Vegas, a five-hour drive away from the wildfire-stricken city of angels. At the convention, the chipmaker showcased some really stupid generative AI gambits like its “Neural Faces” tech, which purports to “cross the uncanny valley” by changing video game characters’ faces — all while eating up a ton of carbon-emission-causing electricity and water, both intrinsically linked to the devastation in LA.
It’d be one thing if LA weren’t literally on fire. But for the company to boast about energy-intensive crap that nobody wants while people lose their homes and lives just one state over is, as Aftermath suggests, feels beyond the pale.
While AI alone can’t be blamed for the blazes that ripped through the Los Angeles area this past week, new findings out of the University of California, Riverside and Caltech — which is itself located just a few miles from the fire-leveled town of Altadena — show how harmful AI has already become.
As those Golden State scientists suggest in a not-yet-peer-reviewed paper, AI is not only drinking up tons of water to cool data centers, but has already begun escalating carbon emissions and water consumption. Though that paper focuses more on AI’s impact on public health, it notes, as so many others have, that the technology is already proving harmful — and that it will only get worse as it grows. Remember, OpenAI CEO Sam Altman is seeking trillions of dollars in infrastructure to power the industry he envisions.
And for every well-intentioned scientist like Fan, there are a dozen others looking to make a quick buck from worthless “innovations” that come with a hefty environmental price tag.
At the end of the day, an Nvidia scientist suggesting that AI can help solve these climate change-fueled fires feels akin (albeit on a smaller scale) to oil companies selling so-called “carbon credits” that purport — dubiously, critics say — to offset greenhouse gas emissions.
Though there are obviously much worse polluters in the world right now than AI companies, it won’t stay that way if the industry continues to grow at its current speed — and if that happens, there’s little doubt that more Earth will be scorched.
More on AI and the environment: People Think the Hollywood Sign Is on Fire Because of AI Slop