Source: Depositphotos
As artificial intelligence permeates everything, its voracious appetite for energy is reshaping the way global power is sourced and traded. The surging demand for energy-intensive GPUs that feed the AI beast risks straining national grids and testing sustainability goals over the coming decade.
Not so long ago, it was Proof-of-Work cryptocurrencies such as Bitcoin and Ethereum that were attracting flak for their energy footprint. While Bitcoin mining remains a power-hungry process, Ethereum has moved on to greener pastures following its transition to Proof-of-Stake. But no sooner had the environmental lobby gotten over its critique of crypto, it’s found another adjacent industry to attack in the form of AI.And it’s hard to dispute the facts as they stand: AI has one hell of an appetite for energy. From training massive language models to powering real-time inference in data centers, AI’s energy consumption is skyrocketing. All transformative technologies, to be fair, entail a massive spike in production – a trend that can be traced as far back to the Industrial Revolution of the 18th century. It’s the price of progress.Nevertheless, the rate at which AI’s energy footprint has increased poses a pressing question: how can the world meet this energy need sustainably and reliably? The answer lies in a blend of technological foresight and strategic energy investment. And one energy source in particular has the potential to meet that need. It is, unsurprisingly, nuclear. What’s perhaps more surprising is the opportunity it presents shrewd investors to load up on U3O8 – yellowcake used in uranium production – without needing a geiger counter and a lead suit.
AI Is Coming for Your Energy
We commonly envision AI as an energy saver, freeing us from the drudgery of planning meals, preparing reports, and designing school newsletters. But we rarely consider that for everything AI gives back to us in free time, it consumes in other ways. GPUs, such as those in Nvidia’s H100, draw 700 watts each, and a single data center can house thousands. Do the math, as they say.
While AI’s relationship with energy is complex, regions where there is a high concentration of data centers place a significant impact on the local grid. Data centers are typically clustered together for efficiency reasons, which demands reliable energy infrastructure that can keep up. Renewables such as wind and tide aren’t cut out for this sort of work. Nuclear, on the other hand, is a perfect fit with its ability to crank up capacity at the push of a button.
As the U.S. seeks to recenter itself as a tech hub, with a particular focus on AI and crypto, it’s safe to say its energy footprint is only going to climb. Particularly as new chip manufacturing plants come onstream to keep the AI juggernaut ticking over. Today’s energy mix, dominated by fossil fuels, with nuclear a distant third behind renewables, faces a reckoning.
Solar and wind, while growing, are intermittent, leaving gaps that can’t sustain AI’s round-the-clock demand. Battery storage, also advancing, can’t yet bridge the lulls for data centers needing constant power. Fossil fuels, meanwhile, clash with net-zero targets, with this tension pitting AI’s promise against sustainability. Which is why the demand for a reliable and scalable energy solution brings us back to nuclear.
Uranium Demand Ramps Up
Given its zero-emission design and predictable output, it’s no wonder nuclear energy is being mentioned in the same breath as AI. The future technology of the 1950s is now the power source for the future tech of the 2020s. As a result, uranium – the primary fuel for nuclear reactors – is seeing renewed interest, particularly as countries pursue net-zero goals.
One of the interesting side-effects of uranium’s increased demand is that it’s presented an opportunity for astute investors to capitalize on rising global energy consumption, exacerbated by an increase in AI. Enterprising RWA platforms have latched onto this, with the likes of xU308 enabling traders to acquire tokenized uranium. Operating on the Tezos blockchain, xU308 presents a compelling case for uranium investment, noting that it has delivered more than 3x the return of the S&P 500 in the last five years.
1/4 ⚛️ Built on @etherlink and powered by @Tezos, is the world’s first decentralized marketplace for trading tokenized physical uranium. This platform makes it possible for retail investors to access uranium without having to spend more than 5 tez. 🧪 pic.twitter.com/vVc1i4bAUB
— Tezos (@tezos) December 5, 2024
As more RWAs arrive onchain, including many associated with the AI economy – such as NVIDIA stock and tokenized GPU compute – uranium presents an asymmetric bet on rising demand for artificial intelligence. For investors seeking portfolio diversification coupled with a future-proofed bet on growth in global energy demand, uranium looks like a reliable choice. Lowly correlated with other commodities, it appears largely immune from geopolitical crises and tariff wars.
In the long-term, anticipated improvements in solar, wind, and battery technology may allow renewables to take over from fossil fuels. But for the next 5-10 years, it’s clear that there’s a gap to be filled between fossil fuels, with their dirty but effective capacity, and renewables with their capricious promise of a greener world. Into that gap nuclear effectively slots, providing the short to medium-term answer to the world’s energy demands.
Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.