Propelled by technological advancement, low interest rates, broad and frenzied investment appetite, and a laissez-faire regulatory environment, the path of human development is on the verge of material change even though it may not be abundantly clear now. Though this could describe the nascent phase of artificial intelligence (AI), it is also a description of the early days of the U.K. railway boom, which was a crucial component of the industrial revolution of the 19th century.

While railways were the subject of immense investment and profit, they were also the conduit that transformed the broader economy, increased productivity, and accelerated per capita GDP for decades thereafter. With rail’s introduction, agriculture, raw materials and manufactured goods could be transported further and cheaper, speed of commerce and communication increased, consumers had greater tourism choices, regionalism gave way to nationalism, and militaries could deploy faster. Even our concept of time changed. Before trains, every town had its own local time. To keep the trains on schedule, “railway time” was created, which led to the time zones we use today. In short, railways were the engine of the first industrial revolution.

Likewise, electricity, internal combustion engines, and assembly line production drove the second industrial revolution in the late-19th/early-20th centuries. Electronics, computers, and automated production defined the third industrial revolution in the mid- to late-20th century. Experts suggest the current “information age” marks the fourth industrial revolution which has been enabled by the internet and smartphones. AI, while still at an emerging stage, is seen by some as an evolution within the current revolution, and by others as potentially giving way to a fifth. AI has many of the hallmarks of its technological forerunners, and as such, we may be on the cusp of a new wave of economy-wide productivity gains, new applications/products, and improved living standards. Parallel with previous revolutions, disruption, obsolescence, environmental and societal impacts are also inevitable.

Industrial revolutions are synonymous with productivity gains
Productivity is a measure of economic performance and efficiency, indicating how well inputs (like labour or capital) are converted into outputs (products or services). It’s measured by the formula: Productivity = Output / Input, where higher ratios signify better efficiency. Common measures include labour productivity, which is output per hour worked, and multifactor productivity, which considers multiple inputs like labour and capital combined.

Based on data from the International Monetary Fund, economic growth was largely linear and linked to population growth for hundreds of years preceding the 1800s. Indeed, without innovation, output was limited to how much a person could produce with their manual labour. Starting in the late-18th century, and coinciding with the first industrial revolution, GDP per capita accelerated at a “hockey stick” pace and hasn’t stopped as wave upon wave of innovation has driven productivity gains. With global population growth expected to decelerate and peak later this century, a greater share of economic growth will depend on productivity gains.

Laying the Tracks for the AI Future
Technological progress is considered a revolution when it has far-reaching economic, societal, climate, and political implications. Just as the range of outcomes of the first railways or lightbulbs or computers was uncertain, AI’s eventual impact will only be visible in the rearview mirror. Will AI be a disappointment or could it be transformative? Will progress outweigh the disruption that AI causes? Will wealth creation/destruction become more concentrated?

It may be early innings for AI, but the global annual investment cadence already exceeds the total combined spending on the Apollo Program, Manhattan Project, and mapping of the Human Genome (in inflation-adjusted dollars). Like the laying of rail and development of locomotives, much of the AI investment to date has been in infrastructure: hyper-fast computer chips, data centres, and training of large language models. However, the wider promise of AI lies in its applications. We’re already seeing it in action with:

• Netflix and Spotify suggesting what to watch or listen to next
• Siri and Alexa helping with daily tasks
• Banks spotting fraud to protect your money
• Self-driving cars
• New ways to find diseases and speed drug development

As capability and cost structures evolve, adoption and novel applications are expected to flourish. Just as smartphones did not exist at the dawn of the internet, we are convinced there are applications of AI that have yet to be imagined.

AI, Agentic AI, and AGI: What’s the Difference?
The landscape of artificial intelligence is rapidly evolving, with new terminology emerging to describe the growing capabilities of these systems. While often used interchangeably, AI, agentic AI, and artificial general intelligence (AGI) represent distinct concepts with significant differences in their scope and functionality.

Artificial Intelligence (AI)
At its core, AI is a broad field of computer science focused on creating systems that can perform tasks that typically require human intelligence. This encompasses a wide range of technologies and techniques, including machine learning, natural language processing, and computer vision. Algorithms on streaming services, spam filters in your email, and voice assistants on your phone are all examples of what is often referred to as narrow AI or weak AI. While highly effective at their designated function, they lack the ability to apply their knowledge to other tasks.

Agentic AI
Agentic AI is a more recent development that focuses on creating autonomous systems, or “agents,” that can proactively and independently take actions to achieve specific goals. Unlike traditional AI that primarily reacts to human input, agentic AI can perceive its environment, make decisions, and execute tasks with minimal human intervention. Imagine a sophisticated personal assistant that doesn’t just answer your questions but can also make travel arrangements, manage your calendar, and negotiate prices on your behalf. This is the promise of agentic AI. These systems are designed to be goal-oriented and can often break down a complex objective into a series of smaller, actionable steps.

While agentic AI exhibits a higher degree of autonomy than narrow AI, it is still task specific. An agent designed for e-commerce, for example, would not be able to apply its skills to medical diagnosis. However, the development of agentic AI is a significant step towards more capable and independent AI systems.

Artificial General Intelligence (AGI)
Artificial General Intelligence represents a hypothetical future stage of AI development. An AGI would be able to understand, learn, and act across a wide range of tasks similar to humans or even better. AGI combined with robotics could expand into physical tasks.

Characteristics of a true AGI could include:

• Common sense reasoning: the ability to understand and navigate the complexities of the world in a way that is not explicitly programmed
• Creativity and problem-solving: the capacity to generate novel ideas and find solutions to unfamiliar problems
• Self-awareness and consciousness: a subjective understanding of its own existence and mental state (a highly debated and philosophical aspect of AGI)

Though today’s most advanced AI models demonstrate impressive capabilities, they do not possess the generalized intelligence and adaptability that would define AGI.

Innovation has been a driver of long-term equity market performance

Equity markets are subject to cyclical forces such as interest rates, fiscal policy, unemployment rates, and inflation. However, demographics, climate change, rising national debt, and innovation are secular influences that help shape longer-term trends. With much of the third and fourth industrial revolutions centred in the U.S., productivity gains have helped drive meaningful outperformance of the S&P500 Index (total return of 15,258%, 1970-2019, in USD) over the MSCI World Index (total return of 4,110%, 1970-2019, in USD); all data per Bloomberg. This outperformance came despite slower population growth in the U.S (63%, 1970-2019) compared with global population growth of 112% over a similar period. Clearly, innovation matters.

At this early juncture, it’s too early to quantify the impact AI will have on future productivity gains and equity market performance. Just as in the rail boom of the 1840s and the internet boom of the late-1990s, initial rewards of the AI era have accrued to infrastructure builders. Should AI-driven innovation follow historical patterns, the productivity benefits of AI should spread across the broad economy over time and lift many other boats, not just the suppliers of the “picks and shovels”. This means that diversification across sectors is still valuable. It is noteworthy that stock prices of U.K. railways, electric utilities, automakers, internet companies initially outperformed during their respective booms, but later dropped on concerns of overcapacity and excess valuation.

Market resilience amidst headwinds
Equity markets have demonstrated remarkable resilience this year, posting strong gains despite a backdrop of tariff uncertainties and signs of an economic slowdown. This apparent contradiction can be attributed to a forward-looking investor perspective that is prioritizing resilient corporate earnings, prospect of looser monetary and fiscal policy, and transformative technological trends over near-term macroeconomic headwinds.

Underpinning much of the market’s strength is a powerful secular narrative centred on AI. Investors are increasingly looking to the transformative potential of AI to boost productivity, create new markets, and drive corporate profits for years to come. This has led to significant inflows into technology stocks and a willingness to look beyond current valuations.

Geographically, our managed portfolios are overweight the U.S., relatively neutral in Canada, and underweight international. We have taken a balanced approach by building exposure to technology while remaining diversified and cautious of valuation levels. Within our growth-oriented portfolios, our technology and AI exposure is gained through direct positions in industry leaders Alphabet (GOOG) and Microsoft (MSFT), while iShares U.S. Technology ETF (IYW) and iShares S&P/TSX Technology ETF (XIT) provide broad industry exposure, and actively managed Fidelity Global Innovators Fund is focused on innovative and disruptive companies. Our portfolios are overweight in the financial services sector, which we consider attractive for cyclical reasons and potential early beneficiaries of AI implementation.

As always, we encourage you to contact us if you have any questions regarding your portfolio and our market outlook.