Marketplace All-in-One Is Google honest about the environmental impacts of Gemini?
Sep 16, 2025
In this episode, Emma Strubell, an assistant professor at Carnegie Mellon University specializing in the environmental impact of AI, challenges Google's claims about its AI model, Gemini. She argues that energy efficiency stats might not reveal the complete picture, stressing the need for better metrics on resource consumption. The discussion also dives into the broader implications of AI technologies on the environment, emphasizing transparency in energy use, and offers best practices for utilizing AI responsibly.
AI Snips
Chapters
Transcript
Episode notes
Per-Query Metrics Can Mislead
- Google's per-query energy and water numbers can obscure larger environmental impacts when multiplied across many queries.
- Emma Strubell warns that single-query metrics can be misleading without total usage context.
Question The Value Of Each Query
- Evaluate whether each AI use case justifies its energy and resource cost before deploying it widely.
- Prioritize AI applications that genuinely advance climate and societal goals over ad-driven or trivial uses.
Efficiency Gains Aren't The Whole Story
- Efficiency gains per query can be outweighed by fast growth in total queries.
- Emma Strubell notes Google does not disclose overall query volume, which matters for total impact.