Environmental harms from operating LMs

Taxonomy of Risks posed by Language Models

Weidinger et al. (2022)

Source DOI

Sub-category

Risk Domain

6Socioeconomic & Environmental

6.6Environmental harm

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.

"LMs (and AI more broadly) can have an environmental impact at different levels, including: (1) direct impacts from the energy used to train or operate the LM, (2) secondary impacts due to emissions from LM-based applications, (3) system-level impacts as LM-based applications influence human behaviour (e.g. increasing environmental awareness or consumption), and (4) resource impacts on precious metals and other materials required to build hardware on which the computations are run e.g. data centres, chips, or devices. Some evidence exists on (1), but (2) and (3) will likely be more significant for overall CO2 emissions, and harder to measure [96]. (4) may become more significant if LM-based applications lead to more computations being run on mobile devices, increasing overall demand, and is modulated by life-cycles of hardware."(p. 220)

Entity— Who or what caused the harm

Human

Due to a decision or action made by humans

AI system

Due to a decision or action made by an AI system

Other

Due to some other reason or is ambiguous

Intent— Whether the harm was intentional or accidental

Intentional

Due to an expected outcome from pursuing a goal

Unintentional

Due to an unexpected outcome from pursuing a goal

Other

Without clearly specifying the intentionality

Timing— Whether the risk is pre- or post-deployment

Pre-deployment

Occurring before the AI is deployed

Post-deployment

Occurring after the AI model has been trained and deployed

Other

Without a clearly specified time of occurrence

Supporting Evidence (1)

"The wider environmental impact of operating LMs may be sig- nificant, however specific forecasts are missing and emissions will depend on some factors which are currently unknown [96], includ- ing (perhaps most importantly) what types of applications LMs will be integrated into, the anticipated scale and frequency of LM use, and energy cost per prompt. Ultimately, the energy require- ments and associated environmental impact of operating large-scale LMs may be anticipated to also exceed the cost of training them, especially when LMs are used more widely"

Part of Risk area 6: Environmental and Socioeconomic harms