Google says it’s now falling short of its climate-conscious goals. Why? Artificial intelligence.
The tech giant announced this week in its 2024 environmental report that its greenhouse gas emissions have risen nearly 50% since 2019, due in large part to the increasing demands of the data centers and supply chains used to build and power its AI technology.
It’s a significant setback to Google’s aim of achieving zero emissions by 2030. In the report, officials admit to “significant uncertainty” around that goal, due to “the uncertainty around the future environmental impact of AI, which is complex and difficult to predict.”
One has trouble imagining how the situation might improve between now and then. We do not yet have a standardized method for quantifying the total amount of emissions caused by AI use, but we do know that the data centers powering them (which are often reliant upon fossil fuels to function) account for nearly 4% of all global greenhouse gas emissions. For the record, that’s a greater amount than the entire aviation industry. Also, the rate at which we mine for the rare-earth metals needed to make AI data chips may outclip global supply – yet demand is only increasing.
After pumping billions of dollars in this emerging tech, however, there’s little chance of Google turning heel on this particular investment. Especially since the worldwide AI market is poised for astounding growth, with a projected worth of $826.7 billion by 2030 (the same year as Google’s zero-emissions promise).
What’s It All For?
Meanwhile, the strain on our planet is likely to increase by 160% in that same time frame. Such a growing global imposition upon our resources demands an answer to the question: What is it that we gain and achieve through such a collective environmental sacrifice?
To be fair, there are positive applications of AI technology in existence. Some examples: Educators utilize AI to craft lesson plans that better speak to their students’s needs. Image descriptions and other such accommodations help people with disabilities navigate the world with increased ease. Healthcare providers can leverage AI to more efficiently scan for warning signs of disease in patients.
But while those are notable beneficial use cases, they do not negate the harms caused by AI tech. The problems begin with who’s behind the technology – in a surprise to no one, it’s largely white men engineering the product of the future. (The committees pieced together to oversee AI implementation at our biggest companies also tend to be, shall we say, depressingly homogenous.) Though women have found opportunities in this emerging space, a 2021 Deloitte report found that they comprise just 22% of all AI workers.
Indeed, the AI space itself sees the same tired sorts of gendered discrimination found in other parts of the tech world, further excluding women innovators. “Nobody took us seriously, not a single person,” Davar Ardalan, an AI company founder, told The Story Exchange while recalling rejections from 350 investors. “It’s incredibly demeaning.”
The implicit bias of AI’s creators manifests in myriad ways on the consumer-facing side of things. Researchers at Capitol Technology University, a private university outside of Washington, D.C., found numerous examples of AI harming women in their March 2024 study. These includes threats to our job security, with AI candidate-search tools weeding out more women jobseekers while AI technologies automate the administrative and retail jobs women are more likely to hold. And when it comes to our medical care, AI scanning and assessment tools work off of skewed data that minimizes the importance of women’s symptoms.
AI generated images are also rife with sexism, often making assumptions about the professions women can occupy and almost always distorting us to better conform to younger, Eurocentric beauty standards, whether or not a user’s prompt asked for such specifications. Far worse are the fake pornographic images depicting real women in states of undress, crafted without consent – and there is, as of now, no recourse for such an intrusion.
And beyond the biases within the industry and technology itself, it’s worth noting that the climate-change impacts of powering AI are likely to disproportionately harm women, too.
Finding a Way Out
One thing we need, beyond awareness, is regulation and legislation. Of course, it’s a hell of a time to turn toward either concept as a solution, as our democratic process faces unprecedented challenges and oversight agencies reel from a U.S. Supreme Court ruling that will deeply undermine their authority. Yes, that includes those in the tech sector.
But there are a number of bills working their way through state legislatures that provide paths forward. In states like California, Vermont, New York and Oklahoma, for example, bills have been put forth that place the onus on AI developers to weed out discriminatory outcomes generated for users. On the federal level, a bill has been floated that would, if enacted, require the study of AI’s environmental impacts, and mandate plans of action around the findings.
Beyond that, we also need more women involved in the development of AI, insiders say. As tech founder Ardalan noted to us: “There’s a lot of amazing ideas that women have that make AI more empathetic, that make it more human-centered, that bring solutions that really can help bring people together and make really interesting experiences.”
This is about more than staving off soulless Toys ‘R’ Us ads or the (seemingly inherent) goofiness of AI-generated hands. We are ravaging our own planet – a world experts say is on track to becoming “unlivable” – to grow technology that further drains us of what we need, all while it discriminates against women in both its development and implementation.
We need change, and answers. And not from ChatGPT.◼
