The AI revolution: Reshaping industries and investment landscapes

The rapid evolution of artificial intelligence (AI) is no longer confined to theoretical discussions or niche applications—it is fundamentally reshaping global industries, infrastructure, and economic paradigms. In a recent episode of The Angle from T. Rowe Price, Eric Veiel, Head of Global Investments and CIO, sat down with NVIDIA CEO Jensen Huang to explore the transformative potential of AI, its energy implications, and the emergence of new industrial frameworks.

The Evolution of Computing Architecture: Beyond General-Purpose Systems

For decades, computing primarily relied on general-purpose CPUs, a one-toolfits-all akin to using a hammer for every job. This paradigm is being overturned by accelerated computing, where specialised processors like GPUs optimise specific tasks. Huang emphasised that 5% of code represents 99% of compute time—an inefficiency that can potentially be addressed by offloading intensive workloads to parallel processors. This shift has enabled breakthroughs in AI training, reducing energy consumption by orders of magnitude.

The implications are profound: accelerated computing is democratising access to high-performance capabilities. Eric Veiel noted the “shattering” of Moore’s Law, with Huang adding that, taking NVIDIA as an example, its advancements have driven substantial cost reductions in inference every two years—surpassing traditional semiconductor scaling. This efficiency is not merely technical; it is economic. By slashing energy use and costs, AI is transitioning from a research novelty to an industrial necessity.

Energy Efficiency and the Jevons Paradox: A Double-Edged Sword

Energy constraints loom large over AI’s expansion. Huang highlighted that ‘performance per watt is our metric for success,’ stressing NVIDIA’s focus on maximising throughput while minimising power consumption. However, as Veiel observed, efficiency gains often spur demand—a dynamic encapsulated by the Jevons Paradox. While accelerated computing could reduce the USD1 trillion global data center spend, it simultaneously fuels the rise of “AI factories,” a new infrastructure layer dedicated to intelligence production.

Huang explained this duality: ‘On the one hand we revolutionised the way that computing is done and drove energy efficiency. On the other hand, we made computing so cost effective it created something new. And so, a new type of data center emerged. And this new type of data center is called AI factories.’ 

AI factories, he argued, represent a paradigm shift akin to the birth of automotive or telecommunications sectors. These facilities are expected to generate “tokens” of intelligence— data outputs tailored to industries like healthcare, finance, and logistics. For investors, this signals dual opportunities: optimising legacy infrastructure and capitalising on emergent AI-driven ecosystems.

The Rise of Agentic and Physical AI: From Perception to Action

AI’s evolution is unfolding in distinct waves. The first, perception AI, enabled systems to interpret data (e.g., image recognition). The second, generative AI, unlocked content creation. We now enter the third wave: Agentic AI, where according to Huang, systems ‘perceive, reason, and use tools, plan, take action.’ Huang predicts enterprises will deploy digital agents at scale in 2025, with early production of physical AI (robotics) following in 2026.

Early adopters are already testing prototypes. Huang suggested that physical AI is expected to help address global labour shortages, citing a global deficit of 30–50 million workers. He believes the probability of industries like assembly, warehouse logistics, and back of restaurant type work not adopting physical AI is very low.

As AI continues to evolve, its potential role in augmenting human capabilities becomes increasingly apparent. Huang envisions a future where AI agents could enhance productivity across professions, stating, ‘Every software engineer is expected to be augmented by five or six AI agents.’ In Huang’s view, this augmentation is not about replacing jobs but about enabling workers to focus on higher-value tasks, thereby driving economic growth and innovation.

Supply Chain Resilience: Transparency and Redundancy

AI’s scalability hinges on resilient supply chains. Taking NVIDIA as an example, their approach of prioritising transparency and diversified partnerships could offer insight for the broader industry. Huang stressed that ‘resilience starts with diversity and redundancy,’ referencing collaborations with TSMC and efforts to onshore manufacturing. Notably, NVIDIA shares multi-year roadmaps publicly, enabling partners to align investments.

This strategy reflects a shift from secrecy to collaboration. As Huang noted, being transparent can help answers thousands of decisions globally, from chip design to data center construction. ‘This level of communications transparency, up and down the supply chain is really important for resilience.’

The Democratisation of Intelligence: Closing the Social Divide

Beyond economics, AI offers the potential for societal transformation. Huang envisions it as ‘the single greatest technology that could close the social divide,’ noting that prompting—the new programming language—is accessible to all. Unlike traditional coding practiced by a relative few in the global population, AI interaction requires only natural language. This democratisation could amplify productivity across sectors, from education to small enterprises.

Veiel highlighted practical applications, such as portfolio managers using AI for deep research, while Huang shared anecdotes of quantum chemists and meteorologists working towards breakthroughs, and explained how using AI to predict the weather reduced energy use by 10,000x.

Navigating the AI Revolution

The AI industry’s developments are reshaping the economic and technological landscape. The focus on energy efficiency, the emergence of AI factories, and the advent of Agentic AI are driving forces behind this transformation. As the industry continues to evolve, its ability to augment human capabilities and create new economic opportunities will be pivotal in shaping the future.