Sustainability reports from Google and Microsoft indicate they are struggling on climate action, with both companies citing expansion of data centres for AI, which have high energy and water demands, coupled with a scarcity of renewable energy.
Is AI so profitable they are willing to forego environmental goals? Or do they have no control over AI as a force driving the direction of their business?
Google’s response includes optimising AI training time, adopting next-generation computing power and investing in renewable energy. Microsoft emphasises water and energy efficiency measures (leveraging AI), green steel development, purchasing lower-carbon inputs and carbon removal technology. Additionally, Google and Boston Consulting Group claim that AI could mitigate up to 10% of global emissions by 2030.
These approaches are far from progressive: idolising growth without a clear social purpose, speculating on environmental rewards of innovation and overstating the mitigation potential of efficiency and offsets. These merely aim for an increasingly efficient and enhanced capitalist system without questioning whether the system can actually achieve stated sustainability goals.
For instance, one reason for the scarcity of renewable energy is that although it is increasingly cheap to produce and attractive to consumers, it does not represent an attractive investment. Capitalism itself is an obstacle to the vision of green capitalism. The capitalist drive for perpetual growth underlies environmental crises, income and wealth inequality and environmental injustices.
The purpose of big tech must shift from profitability to social and environmental wellbeing. Without these priorities, profitability is merely another form of extraction and organisational resilience is reduced to financial absorption of shocks.
To genuinely provision socially useful technology within planetary and community tolerances for extraction and pollution, big tech companies must dig deeper for effective strategies for the use of AI and AI’s use of resources.
- AI use should focus on human needs, such as healthcare, education and food production, and concentrate on optimising equitable global trade, resource redistribution, fair labour and ethical sourcing.
- AI’s energy and water demands must consider bio-regional limits for regeneration and assume an allocation after community needs are met.
Image by Alina Grubnyak on Unsplash
Heliocene 