Energy demands from data centres and artificial intelligence (AI) could prove to be a little-talked-about, but extremely integral factor in the nuclear power debate.
The subject arose on June 20 when Opposition Leader Peter Dutton highlighted the increasing power needs for increasing AI and data technology on ABC Breakfast’s radio program.
Mr. Dutton spoke about the importance of data centres in propping up regional businesses, as they could offer a lifeline for business and the economy.
“We’ve had in-depth conversations with our local members who know their communities better than anyone about the options available,” he said.
“I see some of the commentary in relationship to Lithgow, for example, where shops are shut up. We have the ability to bring in datasets. AI will play a huge role into the future, but it’s very energy intensive.”
Datasets are, essentially, organised collections of data that can help businesses run successfully.
The information could be used for something as advanced as scientific analysis or something as rudimentary as working out which pairs of shoes customers are most attracted to.
But that data needs power in ever-increasing amounts—especially if it’s going to help save regional towns and cities.
The World Economic Forum reported in April that the power needed to sustain AI’s increasing presence in our lives is doubling about every 100 days.
In less than five years’ time, AI could be using more electricity than the entire country of Iceland used in 2021.
Regional industries function on low costs, and with that comes a need for reliable, affordable power.
“That’s the reason that we’ve looked at the world experience,” Mr. Dutton said.
“And we believe that there’s significant benefit for that community.”
Regardless of the solution, ever-increasing power needs is an issue that will only increase in significance in the future.
By the World Economic Forum’s data, a 10-fold improvement in AI efficiency could cause power needs to surge by up to 10,000 times.
There is some talk of moving towards the use of shared data centres and cloud computing resources to save money on individual sites and even planning larger tasks for cooler months.
The energy required to run AI tasks is already accelerating, with an annual growth rate between 26 and 36 percent.
The question of how to power an ever-evolving world of technology—or anything at all—has become a very hot political topic.
Prime Minister Anthony Albanese told Sky News that Mr. Dutton’s plan to build seven nuclear reactors across the nation if he was elected as Australia’s leader would amount to “economic catastrophe.”
Mr. Albanese criticised the lack of detail around planning, including whether it will feature community planning processes.
“They’ve made an announcement without any substance, no costings, no real timelines, no idea of what form the reactors will take, how big they will be,” he said on June 20.
“This is farcical. Frankly, this is an economic catastrophe. When we look at the cost of nuclear, what we know is that it is the most expensive form of new energy.”
