Warning: This content has not yet been fully revised for this year.
Background
In our current world, there is a large push for sustainability. This includes conserving the environment by preserving natural land, reducing carbon emissions, developing renewable energy sources, eliminating waste, etc. In a study assessing the potential impact of AI on the UN’s 2017 Sustainable Development Goals (which encompasses societal, economic, and environmental outcomes), researchers found that AI could potentially enable 93% of the UN’s environmental targets.
Examples
- Internet of Things (IoT) devices and appliances
- smart and low-carbon cities
- smart power grids
- drones for combating marine pollution
- geographical data processing with satellite imagery
Potential Concerns
- energy-intensive GPUs add to environmental degradation
- control (loss?) over large-scale AI systems
- AI making decisions on energy use for homes and cities
- issues with big data
Provocations
- What is the discerning factor over the trade-off between running highly accurate AI systems and running low energy-intensive AI systems?
- How can AI be used in (smaller?) sustainability efforts without leaving a large carbon footprint?
- What are the ethical design considerations that need to be addressed with large-scale AI systems (i.e. running a power grid)?