Navigating AI’s Environmental Impact: Mistral’s Groundbreaking Audit
Mistral, a French model-maker, is setting a precedent by unveiling the results of its first environmental audit, shedding light on the carbon footprint and water usage of large language models (LLMs). Here’s what you need to know:
-
Key Findings:
- 85.5% of CO2 emissions and 91% of water use stem from model training and inference.
- Each average AI query emits only 1.14 grams of CO2 and consumes 45 milliliters of water.
- Over 18 months, cumulative emissions reached 20.4 ktons CO2, equivalent to 4,500 cars on the road for a year.
-
Comparative Context:
- An AI query’s impact is akin to watching 10 seconds of streaming or spending 10 minutes on an email read by 100 people.
As our thirst for AI grows, understanding its environmental consequences is vital. Let’s discuss and share insights on balancing tech innovation with sustainability!
Engage with this post and join the conversation on AI’s future!
