Andrew was back on BBC Radio, speaking to presenter John Darvall on Radio Bristol to answer the question:
What impact could generative AI have on the environment? Both positives and negatives?
Did you know that asking 20-50 questions on ChatGPT consumes half a litre of water to cool the chips? Or that training a single AI model like GPT-3 required 700,000 litres of clean water, enough to produce 370 BMWs? These aren’t just stats, they’re the hidden environmental costs of AI.
Generative AI is powerful, but it’s also incredibly resource-hungry. The energy demands are so immense that Microsoft has signed a 20-year deal to reopen the dormant Three Mile Island nuclear plant to power its AI ambitions. This isn’t just a tech story—it’s a conversation starter for your next pub chat.
But here’s the twist: while AI has environmental downsides, it also holds immense potential for good. It can optimise energy use, reduce waste, monitor environmental impacts, and even help us design smarter, more sustainable systems.
Imagine AI analysing supply chains to minimise resource consumption or making real-time adjustments in smart buildings to save energy. The possibilities are vast.
The real question is: how do we balance innovation with responsibility?
I was invited onto BBC Radio Bristol to outline both the positive and negative impacts of generative AI have on the environment. You can listen to the discussion here with John Darvall.
We’re at the very beginning of this journey. AI is no longer just a topic for tech conferences, it’s being discussed in homes, offices, and yes, even pubs. But we need to ask better questions and deploy critical thinking to ensure we maximise its benefits while minimising its costs.
AI isn’t inherently good or bad—it’s a tool. How we use it will define its impact on our planet. Let’s make sure we’re asking the right questions and having the right conversations.