I recently attended a Richmond Events forum on digital and AI developments, where the keynote speaker was technology strategist Dex Hunter-Torricke. His talk, AI & the Crossroads for Society, was one of those presentations that stays with you long after the event ends.
One comparison in particular stuck with me.
Even a simple query to an AI system like ChatGPT requires measurable energy – roughly comparable to running a microwave for about an hour.
It’s a striking way of visualising something most of us rarely think about. Multiply that by millions or even billions of AI queries every day and you start to appreciate the scale of the infrastructure behind the technology.
AI may feel effortless when we interact with it, but behind the scenes it relies on vast data centres and enormous amounts of electricity running constantly in the background.
The Hidden Infrastructure Behind AI
Artificial intelligence is often talked about as if it exists purely in the digital world. In reality, it depends on some of the most energy-intensive computing infrastructure ever built.
Globally, data centres already consume around 415 terawatt-hours of electricity each year, roughly 1.5% of total global electricity demand (International Energy Agency).
And demand is rising rapidly. Some projections suggest electricity use from data centres could more than double by 2030, driven largely by generative AI systems (Deloitte Technology Predictions).
To put that into context, data centres in the United States alone now use roughly the same amount of electricity each year as Pakistan (U.S. Energy Information Administration; World Bank electricity consumption data).
In other words, what appears to be a simple digital interaction is increasingly supported by infrastructure on the scale of national energy systems.
It’s Not Just Electricity
Energy isn’t the only resource involved.
Large data centres generate huge amounts of heat, which means cooling systems must run constantly. Many facilities rely on water cooling, and research has estimated that training a large AI model can consume around 700,000 litres of water (University of California Riverside / Microsoft research).
As AI continues to scale, it’s easy to see why conversations about the technology are beginning to overlap with discussions about energy infrastructure and environmental sustainability.
AI Could Turbo-Charge Scientific Discovery
But focusing only on the costs of AI would miss a big part of the story.
One of Dex’s central arguments was that AI could dramatically accelerate scientific discovery. He described the technology as having the potential to “turbo-charge science.”
Researchers are already using AI systems to analyse enormous datasets in areas such as drug discovery, materials science and climate modelling, identifying patterns that might otherwise take years to uncover.
One particularly exciting field is nuclear fusion research – the long-standing goal of producing clean, virtually unlimited energy. AI models are increasingly being used to simulate plasma behaviour and optimise reactor designs, helping scientists move faster toward practical fusion power (Nature Energy; MIT Plasma Science and Fusion Center research).
If breakthroughs like this happen, the same technology currently driving huge energy demand could ultimately help solve the energy challenge itself.
AI Is Already Changing the Way We Work
Alongside scientific progress, AI is also beginning to reshape how many organisations operate.
Generative AI tools can now assist with tasks that once required highly specialised human input – writing, coding, research, data analysis and marketing content among them.
Studies suggest that AI could help with up to 60–70% of workplace tasks, supporting professionals in completing parts of their roles more efficiently (McKinsey Global Institute).
That doesn’t mean jobs suddenly disappear. In many cases these tools act as productivity assistants, helping professionals work faster and freeing up time for more strategic thinking.
But the pace of change means organisations need to think carefully about how human creativity and expertise remain central as these technologies develop.
The Most Disruptive Decade in History?
Dex also suggested we may be entering the most disruptive decade in economic history, as advances in AI, robotics and automation combine to transform productivity across industries.
Some of the examples he shared bring this into sharp focus.
A robotic brick-laying system, for example, can lay around 400 bricks an hour – roughly six times faster than a skilled human bricklayer. Meanwhile, in highly automated electronics factories in Beijing, production lines can assemble around one iPhone every second!
The point isn’t that people disappear from the process.
But the scale and speed at which technology can now produce things is increasing dramatically. And when productivity shifts that quickly, disruption tends to follow.
What This Means for Businesses – Including Us
Like previous technological shifts – from the internet to smartphones – the businesses that learn to use these tools effectively will gain a significant advantage. Those that don’t risk falling behind.
At Hunterlodge, we see AI not as a replacement for creativity or strategic thinking, but as a tool that can help amplify them.
Used well, AI can help us analyse audiences more deeply, explore ideas more quickly and deliver campaigns that are smarter and more effective. It can also free up time for the things that matter most – insight, creativity and human understanding.
The challenge is to embrace the technology thoughtfully, using it to support better decision-making while keeping people firmly at the centre of the process.
Because ultimately the most powerful combination isn’t AI on its own – it’s AI working alongside human creativity and expertise.
The Question Ahead
AI may feel like magic.
But behind every chatbot response sits electricity, computing infrastructure and the collective knowledge created by people.
The real question isn’t simply how powerful AI becomes.
It’s how wisely we choose to build and use it.
And sometimes it takes a simple comparison – like the energy behind a single AI query – to remind us just how significant the choices ahead may be.
To discuss how AI and human creativity can work together to deliver better marketing outcomes, contact Kim: kim.mclellan@hunterlodge.co.uk