MIT AI Risk Repository · Risk Sub-Category · 17.06.01

Environmental harms from operation LMs

Category: Automation, Access and Environmental Harms

Open the MIT Risk Navigator All entries from Weidinger2021

Description

"Large-scale machine learning models, including LMs, have the potential to create significant environmental costs via their energy demands, the associated carbon emissions for training and operating the models, and the demand for fresh water to cool the data centres where computations are run (Mytton, 2021; Patterson et al., 2021)."

From Ethical and social risks of harm from language models (Weidinger2021), as extracted by the MIT AI Risk Repository (CC BY 4.0).

Classification

Domain: 6. Socioeconomic & Environmental
Subdomain: 6.6 Environmental harm
Causal entity: AI
Intent: Unintentional
Timing: Other

Subdomain definition: The development and operation of AI systems causing environmental harm, such as through energy consumption of data centers, or material and carbon footprints associated with AI hardware.

Real-world incidents in this subdomain

5 Jun 2024xAI Allegedly Operates Unpermitted Methane Turbines in Memphis to Power Supercomputer Colossus to Train Grok

Browse all incidents in this subdomain

How other frameworks describe this risk

Natural resource depletion
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Excessive water consumption
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Excessive energy consumption
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Excessive landfill
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Electronic waste
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Carbon emissions
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Biodiversity loss
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)
Pollution
A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

Other entries from Weidinger2021

Discrimination, Exclusion and Toxicity Risk Category · 1.0
Social stereotypes and unfair discrmination Risk Sub-Category · 1.1
Social stereotypes and unfair discrmination Additional evidence
Exclusionary norms Risk Sub-Category · 1.1
Exclusionary norms Additional evidence
Exclusionary norms Additional evidence
Exclusionary norms Additional evidence
Toxic language Risk Sub-Category · 1.2
Lower performance for some languages and social groups Risk Sub-Category · 1.3
Lower performance for some languages and social groups Additional evidence
Information Hazards Risk Category · 2.1
Compromising privacy by leaking private infiormation Risk Sub-Category · 2.1