How to Invest in AI

• Artificial Intelligence (AI) is the latest buzzword, but it’s difficult to define or classify under a single term, or investment exposure for that matter
• More than a chatbot, AI is supported and used by advanced technologies like semiconductors, cloud storage, and autonomous vehicles
• As AI grows more powerful, its reach will expand, making it vital for investors to adopt a broad view and pursue diverse AI exposure
  

The other day I asked Alexa, ”What’s artificial intelligence?” The response:

“I found this on Wikipedia: Artificial intelligence (AI) is intelligence — perceiving, synthesizing, and inferring information — demonstrated by machines, as opposed to intelligence displayed by humans or by other animals. Intelligence encompasses the ability to learn and to reason, to generalize, and to infer meaning.”

Of course, Alexa also could have just as easily said, “I am.” But defining AI as a consumer-facing device used to play music and report the weather would be far too limiting, given how entrenched and intertwined AI is with so many parts of our economy.

For Investors, There’s More than Meets the AI

Investors make the same mistake by focusing on the consumer-oriented ChatGPT from OpenAI or Microsoft’s co-Pilots. This limited view of AI has fueled returns of some of the more well-known AI-connected stocks, and has led to the narrow market leadership we’ve seen so far this year.

The truth is, AI is so much more than ChatGPT and GPUs — it’s supported and used by advanced technologies, from semiconductors and cloud storage to medical devices and autonomous vehicles. This has significant implications for investors. If you want to pursue the full value of this vast ecosystem, you’ll need to take an increasingly broad view of AI.

AI Is a Force for Transformation, Not a Sector

Think of AI not as a sector or industry, but as a key catalyst of transformational technological changes that are impacting our society. AI is one of the technological pillars (alongside hyper-connectivity, supercomputing, and biological innovations) that allows economies to create more value with less inputs in order to propel economic growth. In fact, Goldman Sachs recently forecast that the use and application of AI could boost global GDP growth by $7 trillion (or 7%) by 2030, more than the current output of Germany and the UK combined.¹

As a productivity tool that fuels innovation, AI also requires advanced inputs. In a recent report, Barclays separated these mutually reinforcing functions and technologies into part of a broader AI “stack” — or an AI value chain with building blocks stacked atop one another.²

What does the AI stack include?

• Components: Semiconductors generating the massive computing power AI needs
• Cloud services: Cloud service providers offering data storage and computing performance at a massive scale
• Researchers and labs: Research focused on developing the AI large language models upon which end-user applications are built
• Applications: Users of AI technology who perform a function or service

With this framework, AI is no longer a sector or industry or product — it’s a change agent.

AI ETFs Lack Consistency, Supporting Broader View

The question investors should be asking is, “Which companies are likely to benefit from this change agent?” Just as there isn’t a single definition of AI, in an “AI stack” world, there is likely no “pure AI” company.

After all, cloud computing firms do more than cater to AI uses cases. The same is true for semiconductors, which have a wide array of applications. And, even if Microsoft becomes synonymous with generative AI given its planned Office copilot feature,³ the company will still be so much more than an AI firm as a result of its diversified operations.

The lack of agreement on what constitutes a pure AI company shows up in the holdings of ETFs that target AI — there is little to no overlap between them.

Bottom line for investors: Targeting AI purity seems impossible.

But more importantly, investors who pursue AI purity do so at the expense of a potentially more beneficial approach: Owning the full stack of building blocks that underpin the broader technological change being driven by AI.

Seek Broader AI Exposure

AI is the throughline to which many transformational innovations can be traced back. And it will increasingly be so, as it continues to reshape the way we work, play, travel, communicate, and learn.

The best way to invest in AI, then, may be to own every sub-component of the vast ecosystem directly impacted by AI, either through the development of AI or the rapid adoption of AI technology.

Avoiding AI FOMO When the Headlines Hit

AI is exciting, there’s no doubt. For many investors, it’s a headline-making, FOMO-inducing trend. While the buzz about AI is real and hard to ignore, the truth is that its identifiable economic impact remains a futuristic forecast for now.

Economists and consultants continue to debate AI growth rates, even as AI continues to evolve. In the meantime, the range of outcomes for how AI might fuel productivity gains and profits for companies, either as users or supporters of the technology, only gets wider.

Pinpointing pure AI exposures, as desirable as that may sound, looks to be a hopeless task. Instead, I’d encourage you think about AI in broader terms. Consider casting a wider AI net by focusing on the entire AI stack as opposed to individual applications or pure AI exposures.

Taking a more holistic view of the impact AI is likely to have on our economy in the years and decades ahead may help investors more fully capture the true potential of this technology as a transformational driver of growth across an incredibly broad range of sectors and industries.