What will drive the advancement of AI?

Certainly! Here’s the revised content with neutral branding:

—

As I discovered while continuing that line of reporting, building new nuclear plants isn’t straightforward or quick. As my colleague David Rotman outlines in his story, the AI boom might end up relying on another energy source: fossil fuels. So what’s going to power AI? Let’s explore this topic.

When we started discussing this significant project on AI and energy demand, we had numerous conversations about what to include. From the outset, the climate team focused on examining exactly what will provide the electricity needed to run data centers powering AI models. As we mentioned in the main story:

“A data center humming away isn’t necessarily a bad thing. If all data centers were hooked up to solar panels and ran only when the sun was shining, the world would be talking a lot less about AI’s energy consumption.”

However, many AI data centers need to be available constantly. Those used to train models can arguably be more responsive to the changing availability of renewables, since that work can occur in bursts, any time. Once a model is being accessed by the public, though, there needs to be computing power ready all the time. For instance, it’s likely that large tech companies would prefer their services to be available at all hours, rather than only during the day.

Solar and wind power do not seem to align well with much of the AI electricity demand unless paired with energy storage, which increases costs. On the other hand, nuclear power plants typically operate continuously, delivering a steady source of power to the grid.

As you might expect, it can take a considerable amount of time to get a nuclear power plant operational.

Large tech companies can support plans to reopen shuttered plants or extend the operating lifetimes of existing facilities. Additionally, some current plants can make small upgrades to enhance their output. For example, there are plans to upgrade the Columbia Generating Station in eastern Washington, potentially increasing its output by 162 megawatts, which constitutes over 10% of the plant’s current capacity.

Nevertheless, these efforts are unlikely to sufficiently meet the demand that major tech companies anticipate in the future.