Editor’s Note: All opinion section content reflects the views of the individual author only and does not represent a stance taken by The Collegian or its editorial board.
The dialogue surrounding artificial intelligence has seen its fair share of negative discourse, rampant with heavy concerns about continued AI development. These concerns are incredibly justifiable; from killing academia and slaughtering the job industry to exploiting privacy, AI seems to deserve a fate of mass abolishment.
But I’d like to challenge this narrative for a moment. What if the pros of AI are equal to — or even outweigh — the cons? Herein lies a point of contention: There is also an equally important narrative that shows AI’s pivotal, almost critical, role in research.
Global health crises have been at the forefront. As a primary example, environmental public health scientists have scrambled to research the severity and solutions to the air pollution crisis. With its standing as a primary global public health threat that claims 7 million premature deaths annually, the problem demands efficiency. And, in truth, we struggle to keep up.
AI, however, specifically machine learning, has proven incredibly valuable in addressing this issue. It has accelerated research methodologies through data processing, prediction, optimization and real-time monitoring.
Beyond air pollution, AI has also tackled issues like water pollution and remediation, soil pollution and toxicology with peak efficiency. It effectively measures material performance, including how effective certain materials are at remediation, global distribution and the removal of pollutants like methane, as well as the populations affected.
But that’s all just scratching the surface. AI can even cut decision-making time by over 60% purely due to its efficiency. It accelerates the pace toward addressing critical health issues that traditional methods couldn’t touch before.
“I believe it is naive to deny the benefit that AI can faithfully serve.”
However, this poses a massive dichotomy within the environmental public health field. AI is good at optimization, yet it contributes to the very problems it’s researching.
A 2024 study found fossil fuel emissions from electricity generation for AI data centers could cause around 600,000 cases of asthma symptoms and 1,300 premature deaths in 2028. It also noted that U.S. data centers alone currently cost several billion dollars in public health, with a predicted increase to over $20 billion by 2030. Further, generative AI training alone can produce more air pollutants than 10,000 LA-NYC round trips by car.
The newfound methods of AI integration into research are not exempt from these impacts. To tackle something as pervasive as air pollution, an extensive need for AI training, development and data integration is required. This means significantly more resources, and that, in turn, means an even greater indirect impact on public health.
Where, then, do we draw the line? How can we compare and contrast progression versus long-term regression? I don’t know if we as humans are capable of drawing the line with AI usage. Change is possible, but it is entirely dependent on the interests of the people in power. Even if the public wants that change, it may be relegated to an afterthought.
There must be a balance struck between optimization and realism. The dialogue seems to swing on extremist ends; many demand either to undergo heavy AI abolishment, or they press the gas pedal on development.
I believe it is naive to deny the benefit that AI can faithfully serve. However, there must be rationality in its development and regulation; the good in AI can only be strengthened if we imbue it with humanism. Our efforts must shift to resolving its existing issues while realizing that, ultimately, responsibility will not compromise our progress.
Reach Carmel Pan at letters@collegian.com or on social media @RMCollegian.
