Eric Schmidt made some bold statements at the recent AI summit in Washington, DC. The ex Google CEO argued that we shouldn’t limit AI’s energy consumption out of concern for climate change, even suggesting that we focus on AI ahead of climate change as advancing AI will eventually solve the climate crisis.
Schmidt’s ‘AI versus climate change’ comments obviously align with his major business investments in AI companies and his belief in AI’s potential is understandable. AI is already proving valuable in so many areas of our lives and, as far as the climate is concerned, specifically in areas like energy and process optimisation, climate modelling such as predicting droughts in Africa, data analysis and supporting smarter decision-making. There’s no doubt AI will be a massive help in the fight against climate change, but framing AI as a solution that should take precedence over climate action is misguided.
AI is part of the problem
AI can help us tackle environmental challenges, but it is not a silver bullet. It cannot address the root causes of climate change—like industrial emissions, deforestation, and overconsumption without broader systemic and policy changes. With regards to overconsumption, people will buy more stuff when prompted by the AI driven algorithms of Amazon and other ecommerce platforms.
In addition, AI itself is energy intensive. Large-scale data centres and AI model training consume significant energy, contributing to the carbon footprint. For example, doubling AI data centres, as Schmidt suggests, could lead to a sharp rise in global emissions. According to the Climate Action against Disinformation coalition, this could result in an 80% increase in global carbon emissions.
Additional research suggests that AI use in data centres could use as much electricity as a small country such as the Netherlands or Sweden by 2027.
In the US, there is already evidence that the life of coal-fired power plants is being prolonged to meet the rising energy demands of AI. Much of this increased energy demand comes from the added complexity of AI operations – generating AI queries could require as much as 10 times the computing power as a regular online search. Training ChatGPT, the OpenAI system, can use as much energy as 120 US households over the course of a year, the report claims.
If AI is allowed to proliferate at this rate without action being taken to protect the environment, it would be unequivocally disastrous for the world.
Big Tech’s nuclear plans
Schmidt’s former company Google, along with other Big Tech including Amazon and Microsoft, are fast exploring options to accommodate AI’s vast energy needs and have landed on advanced nuclear power as the best option. Although this progressive energy source is being touted as safer and more sustainable than traditional nuclear power plants, it also has its disadvantages. Nuclear energy may be clean from carbon emissions, but it uses highly toxic chemicals and has the potential to be utilised in weapons, which has led some environmentalists to take a strong stance against it.
At 51toCarbonZero, we can see the benefits of nuclear – it’s more reliable than solar or wind, it can be produced onshore meaning less reliance on overseas deals, and it’s relatively cheap in comparison to other energy options. However, nuclear power should not be seen as the future of energy; it has its place, but only as part of a balanced energy economy alongside green energy sources. The drive to power AI must not ignore the investment, development and improvement of green energy technology. And this is what we fear Big Tech is overlooking.
Ignoring climate crisis will affect the vulnerable
Prioritising AI advancement and delaying climate action has broad repercussions, and it’s essential to consider the significant societal impact of postponing climate action. Ignoring the urgent need to cut emissions now would disproportionately affect vulnerable populations and low-income countries, many of which are already facing severe climate impacts. Expecting AI to eventually solve these issues overlooks the moral imperative to protect these communities today.
While AI’s potential to contribute to sustainability is significant, prioritising its evolution over immediate climate action could increase risks to humanity. As Schmidt himself has noted, AI can pose an “existential risk” if left unregulated. A measured approach to AI development is therefore essential, but the climate crisis poses an even greater existential threat, demanding urgent action now.
In summary, we view AI as a super-powerful ‘tool’ in the fight against climate change, but it cannot replace the multifaceted approach needed to solve the climate crises. Climate action requires not only technological innovation and process optimisation but also policy reform, economic restructuring, behavioural change and international cooperation. AI is a critical part of the solution, probably contributing in ways we cannot yet imagine, but it cannot take precedence over the climate crisis that demands our immediate attention.
