The productivity potential 

Paul Krugman once noted, "Productivity isn’t everything, but in the long run it is almost everything." By increasing productivity, the U.S. can improve its standard of living, reduce debt, create jobs in new industries and lower costs. With AI poised to become the next “general purpose technology”, are we on the cusp of a new productivity boom?

In our 2023 whitepaper “The transformative power of generative AI”, we explored scenarios of AI adoption that could accelerate productivity growth to between 1.0 and 4.0% annually over the next decade. Such a boost would be significant, though not without precedent.

Signs of an upswing may already be emerging. Despite past technological advancements failing to sustain productivity growth—averaging just 1.4% over the last decade—productivity has been on the rise since late 2022. While it may be too early to attribute this to AI, the timing coincides with increased capital expenditures and AI adoption, suggesting potential for future growth. As a global tech leader, the U.S. is investing heavily in AI, and a modernized capital stock can lead to substantial efficiency and quality improvements over time.

This has important implications for investor return assumptions as well. In our 2025 Long-term Capital Market Assumptions, we've incorporated a 0.20% AI-driven boost to our annual long-term growth forecasts, despite the early stage of AI adoption. This boost, combined with capital investment, helps offset valuation pressures on long-term equity returns. While it may underestimate AI’s full economic impact, which we plan to reassess as more evidence emerges, it is already enhancing our expectations for economic growth and corporate earnings potential in the coming years.

AI presents meaningful productivity upside

Job transformation or displacement?

AI is posed to significantly impact the workplace, prompting both anticipation and concern. The question of whether robots will "take our jobs" is a valid concern, echoing fears that have surfaced throughout history. While technology has killed all sorts of industries and jobs over time, it has also created new ones, supporting an economy today that is at full employment. Could this time be different?

Unlike past technological advancements that automated physical tasks, AI can perform tasks requiring human intelligence, raising questions about its effect on professional and office jobs. Research from the University of Pennsylvania and OpenAI suggests that up to 80% of the U.S. workforce could see at least 10% of their tasks affected by generative AI, with greater impact on higher-paying roles.

Varying degrees of exposure:

• Highly Exposed Roles: Jobs involving repetitive writing, data processing and programming, such as legal documentation, administrative support, customer service, mid-level programming, accounting and financial analysis.
• Moderately Exposed Roles: High-skilled professions in STEM and healthcare are not immune to automation, but likely benefit from automating lower value tasks, enhancing productivity.
• Less Exposed Roles: Jobs involving physical labor or a strong human element, such as construction and childcare, are less likely to be automated.

In our view, AI will more likely augment rather than replace human capabilities in most jobs. While AI is impressive, many areas will still benefit or require human oversight and feedback for some time. As repetitive and time-consuming “grunt work” becomes automated, workers can spend more of their time on higher complexity tasks, meaningful critical-thinking or creative endeavors, expanding their overall output potential.

Just as ATMs transformed rather than eliminated bank teller roles, the future will likely see significant job transformation as we adapt to working alongside AI. Indeed, according to the World Economic Forum's Future of Jobs Report 2025, the most in-demand skills for the next five years involve technological literacy and socio-emotional skills like curiosity, creativity, and flexibility. 

AI will impact most jobs, but jobs likely transform rather than disappear

Mega-cap tech performance

The rise of AI has become a major force in global stock markets. In the U.S., where tech leadership is most prominent, gains have been concentrated among the "Magnificent 7"—Nvidia, Microsoft, Apple, Alphabet, Meta, Amazon and Tesla—which collectively surged over 100% from early 2023 to the end of 2024.

Unlike the Dot-Com bust, characterized by irrational exuberance and stock prices far exceeding intrinsic values, the current AI boom is driven by substantial upward revisions in earnings forecasts for stocks with significant AI exposure. The Magnificent 7 experienced earnings growth of 31% and 34% over the past two years, while the S&P 500, excluding these giants, saw earnings contract by 4% in 2023, followed by a modest 3% increase in 2024. Today's tech leaders differ significantly from the startup explosion of the Dot-Com era; they are large, established, and highly profitable incumbents, well-positioned to make the substantial capital investments necessary to remain competitive in the AI space.

However, the Magnificent 7's valuations have risen to approximately 30 times forward earnings, and despite a robust earnings outlook, these companies face increased scrutiny. Additionally, the high index concentration has left many investors heavily overweight in U.S. tech, exposing them to potential repricing risks.

We anticipate that the elevated valuations of mega-cap U.S. tech companies will continue to be justified by their quality, cash flow generation, and growth potential. However, the opportunity for upside appears more limited in 2025. While leading tech firms boast a strong earnings growth outlook, investors have become "tiger parents"—quick to reprimand disappointments and quieter in applauding successes. As such, investors should consider diversifying their portfolios to mitigate concentration risk and take advantage of emerging opportunities across the AI value chain. 

The AI Boom does not look like the Dot-Com Bust

The tech-driven capex boom

The dawn of AI has also unleashed a new capex cycle focused on developing AI infrastructure. Leading this spending surge are the hyperscalers—large-scale cloud service providers like Microsoft, Amazon and Google—uniquely positioned to manage and host AI workloads.

In 2024, these major players collectively increased their capital expenditures by over 50%, reaching approximately $200 billion. Our Equities Research Team anticipates that they will invest over $1 trillion in the next five years¹. During their Q4 2024 earnings calls, Amazon projected a $100 billion capex for 2025, Meta is budgeting $60-65 billion (about 33% of its revenue), Google plans to spend $75 billion, and a partnership between OpenAI, Oracle and Softbank aims to invest up to $500 billion in AI infrastructure over the next five years.

These investments are supported by substantial cash flows, unlike previous capex cycles that depended heavily on debt financing. However, such massive spending carries the risk of capital misallocation or unmet expectations. As Bill Gates once put it, people often overestimate what will happen in the next two years and underestimate what will happen in 10. For investors, timing and return on investment will bear significance in the race towards profitable AI adoption by the hyperscalers.

¹ 2025 Long-Term Capital Market Assumptions.

An AI-driven capex boom is underway

The AI Value Chain

Capex has been targeted towards the key infrastructure assets needed to support AI development and deployment. These investments include:

1.       High-Performance Computing: Investing in powerful GPUs and specialized AI chips to accelerate machine learning and deep learning processes. High-speed memory, large-capacity storage solutions and high-bandwidth networking equipment are also needed to facilitate data processing and the fast transfer between data centers and end-users.

2.       Data Centers: Expanding and upgrading data centers to handle the increased computational demands of AI workloads. This includes building new facilities and enhancing existing ones with advanced cooling systems and energy-efficient technologies.

3.       Power infrastructure: Powering up AI is going to require an “all hands on deck” approach.

a.       Investments in renewable energy sources like solar and wind, and energy-efficient technologies such as advanced cooling systems, to power data centers sustainably are likely to continue. Enhancing grid connections and deploying battery storage systems can also help ensure reliable energy supply and optimize distribution.

b.       Localized Energy Systems: Investments in microgrids are growing, particularly in areas prone to natural disasters or with unreliable grid access. These systems can operate independently from the grid and incorporate renewable energy and storage.

c.       Traditional power sources, such as natural gas or backup diesel generators, are also being explored to fill gaps and ensure consistent power availability during peak demand or when renewable sources are insufficient.

4.       AI Platforms and Services: The hyperscalers and several other companies are major players in developing and expanding AI platforms and services for enterprise use. Examples include:

a.       AI model training and deployment tools: Microsoft Azure AI, AWS and Google Cloud AI provide platforms for building, training, and deploying machine learning models at scale and tools for training or deploying AI models in company pipelines.

b.       Specialized AI services: Smaller companies and startups that leverage AI can offer specialized services to specific industry needs, such as predictive analysis for supply chain optimization, crop demand forecasting, medical imaging and talent acquisition.

5.       Security and Compliance: As organizations increasingly prioritize data protection and regulatory adherence, investments towards cybersecurity, fraud detection and prevention and data privacy have become increasing important and diverse.

The capex boom is broadening the range of AI beneficiaries, presenting investors with diverse opportunities to invest in AI. One company’s capex becomes another company’s revenue, and the AI infrastructure value chain has revealed beneficiaries across both value and growth sectors, spanning small-, mid- and large-cap companies globally. 

Investment is flowing across the AI Value Chain

 Investing principles for “the next big thing”

When investing in the “next big thing”, investors should be mindful of a few key principles.

• You don’t get paid in innovation, you get paid in dollars. The Dot-com bubble was driven by a rush to capitalize on technologies without understanding their applications or commercialization. Even the most revolutionary technology can be a poor investment if companies cannot extract profits from it. Industry structure, competitive positioning, regulatory hurdles and economies of scale can all have significant bearing on profitability.
• Stock picking is hard and concentrated stock positions have considerable risk. More than 40% of all companies that were ever in the Russell 3000 Index from 1980 to 2020 experienced a “catastrophic stock price loss”, where they plunged at least 70% and never recovered. You can make a fortune if you pick the right stock, but only about 10% of all stocks since 1980 have met the definition of “mega winners”, whereas a diversified portfolio would have outperformed a concentrated single stock position 2/3rds of the time¹.
• Take a portfolio approach to thematic exposure. Most investors need an allocation to growth stocks to reach their long-term goals, but too much exposure can lead to a precarious amount of risk in portfolios. Most secular trends like AI also create opportunities across various sectors, factors, company sizes and geographies. Investors should consider how the inclusion or exclusion of AI-linked companies can impact overall portfolio tilts and exposures.

¹ Michael Cembalest, Eye of the Market, “The Agony and the Ecstasy”