Can the AI Boom Harm Our Planet and Society?

Microsoft CTO advisor Leslie Miley recently delivered a talk on AI and sustainability, highlighting that the intense computational demands of artificial intelligence require larger data centers, leading to higher CO₂ emissions and environmental damage.

She compared the impact of ChatGPT to the rise of the World Wide Web in the late 1990s and noted that many people liken its influence to that of the iPhone, describing it as a transformative technology that changes where people gather.

According to Miley, few consider AI’s environmental and social consequences. Training AI models is computationally intensive; OpenAI reported in 2018 that the compute required per model is increasing roughly tenfold each year, driven by custom hardware such as GPUs and TPUs and a relentless push by R&D teams to use more chips.

Meta has halted work on European data centers and is repurposing them for AI workloads, because large‑scale AI tasks need a different type of data center than those built for ordinary online services.

New data centers can require 10 million to 19 million liters of water per day for cooling, exacerbating water shortages in nearby communities.

Miley likened the AI revolution to the massive interstate highway program of the 1950s under President Eisenhower, which, while transformative, imposed costs on local communities. She warned that AI will “destroy the environment, devastate communities, and affect children,” citing the relocation of a historic Black cemetery for a Microsoft data center in Virginia as an example.

She also warned about AI bias, noting that AI can not only inherit but amplify societal prejudices. As an illustration, she referenced the fashion brand Levi using AI‑generated models to simulate diversity, resulting in a shallow, culturally appropriative portrayal.

Miley criticized cryptocurrency, stating that early investments revived idle coal‑fired power plants, funneling billions into a sector rife with fraud, and praised China’s ban on mining as a correct move.

She emphasized the responsibility to ensure that technological demand does not compromise the world, urging developers to adopt smaller models and leverage tools such as Azure’s emissions dashboard and AWS’s carbon‑footprint tracking to mitigate impact.