The AI boom is exhausting our power grids
The impact is greater than meets the eye.
The generative AI boom is exhausting the power grid. And tech firms are scrambling for a miracle solution.
We know the AI boom is driving a surge in electricity consumption globally. But by how much?
Exhausting the power grid
According to a report by the 𝘞𝘢𝘴𝘩𝘪𝘯𝘨𝘵𝘰𝘯 𝘗𝘰𝘴𝘵, AI is badly skewing power grids in the US, at least.
Specifically, AI is:
- Delaying the retirement of some coal-fired plants.
- Seeing tech giants snap up finite renewable energy.
- Growing demand backfilled with fossil fuel expansions.
Even as tech giants boost of using 100% renewables, fossil fuel plants are being expanded to stabilise the power grid for data centers.
The reason is simple: Data centres run around-the-clock, but solar and wind are intermittent. In lieu of hugely expensive battery storage systems, fossil plants are often used to fill the power gap.
Other energy sources
The scramble for more power to get ahead in AI is also fuelling research in other sources of power. Could AI save the day by accelerating innovation?
- Next-gen nuclear plants: Small nuclear plants that are touted as safer than previous generations of nuclear facilities could be built right beside data centres to power them.
- Geothermal power: I wrote previously that the 3.7GW of geothermal power in the US comes from 61 different plants. However, the race for AI is seeing much larger geothermal power plants being designed.
- Fusion: Microsoft is apparently bankrolling a startup in fusion energy. However, some have criticised predictions of commercial fusion by 2030 or 2035 call it "hype" at this point.
What do you think?
Will our thirst for more power accelerate innovation, or will it be the death knell to meeting sustainability targets?
