How AI Will Power the Next Industrial Revolution

The next industrial revolution?

It is our belief that artificial intelligence (AI)—computer systems so powerful that they can perform tasks that had previously required human intelligence—will power the next industrial revolution. Given the recent and rapid surge in interest and news coverage of AI, this is likely not the first time you’ve heard such a bold statement. In our view, however, it isn’t hyperbole. Rather, it is a thesis that is justified by a close examination of AI’s development history, established capabilities, and trajectory.

In order to comprehend the magnitude of the societal changes promised by AI, it is worth revisiting another era of immense transformation: the turn of the twentieth century, which saw the introduction of several important technological innovations, including electricity, the automobile, and the telephone. That period showed how technology could fundamentally alter the world in which we live and work—and create billions in shareholder value in the process.

Nearly a century later came the digital revolution. Analysts predicted that communication innovations like the internet and mobile phones would change the world. They did, of course—but as the Dot-com bubble demonstrated to investors, this kind of growth doesn’t necessarily happen in a straight line.

AI is next in that line of world-altering megatrends. Let us explain.

Why AI, and why now?

As any science-fiction fan will tell you, the concept of AI has existed for decades. What happened, then, to propel it to its current status as arguably the most-hyped and fastest-adopted technology of all time?

ChatGPT, which took the world by storm in late 2022, was not the start of AI but rather the culmination of years of mostly off-the-radar research. Prior to the 2010s, progress was slow because of a lack of data and computing power. The breakthrough came in 2012, when researchers at the University of Toronto led by Geoffrey Hinton—now sometimes known as the ‘Godfather of AI’—used a technique called Deep Learning, which simulates how the human brain works by using math functions in place of neurons, to improve image recognition algorithms. Suppose, for instance, you input an image of a cat. Before 2012, computers could only accurately identify it as a cat about 75% of the time, compared to 95% accuracy for human beings. After this breakthrough, however, the accuracy skyrocketed to 96%—even better than humans. The proverbial Rubicon had been crossed, and for their efforts, Hinton and some of his colleagues won the Turing Award, which is considered the Nobel Prize of the computing world.

Image recognition is one thing. Language processing is something else. In order to generate passable, human-esque language, hundreds of billions of parameters (math functions) were needed, compared to the hundreds of millions available only a decade earlier. When that bar was finally cleared, the result was ChatGPT. Suddenly, AI was capable of doing things that were once exclusively the domain of humans—writing essays, generating useful computer code, and even passing a Wharton MBA exam. ChatGPT quickly became the fastest-adopted technology in history, with 100 million users in only two months. (This mark has since been surpassed only by Meta’s Threads microblogging app, which benefitted from users’ ability to sign up with their existing Instagram accounts.)

There’s no question that the tone of discussions about AI has changed over the past several months. We’ve been attending tech-related conferences for years, and prior to 2023, cryptocurrency was the hottest topic. Not so anymore. At a conference we attended in June, every single presentation mentioned some sort of application of AI. It’s already clear that 2023 will be seen by history as artificial intelligence technology’s debutante ball, and understandably, investors are asking—what is this, and how can we get exposure to it?

Source: Visual Capitalist.

The significance of AI for investors

Simply put, the rise of AI will likely generate billions of dollars in market capitalization for existing companies over the next decade, as well as new and significant initial public offerings (IPOs). Through the BMO Global Innovators Fund, BMO Global Asset Management intends to be a leading investor in that AI revolution.

This is a tectonic shift in computing. In 2007, Apple’s introduction of the iPhone ushered in a decade of infrastructure-building, including data centres, mobile infrastructure, and the cloud. The iPhone drove the wider adoption of 4G broadband technology, and its App Store led to the birth or growth of trillions of dollars’ worth of sharing economy companies, including Amazon, Uber, Airbnb, Meta, and Netflix. The 2020s, similarly, are poised to be the breakout decade for AI, leveraging infrastructure that is already in place. As with the sharing economy, investment opportunities start with hardware companies like Nvidia and AMD, which have already seen their valuations skyrocket. Next will come the cloud computing companies, and then the software and applications that will be the predominant foundations of this boom; ChatGPT has already given us a modest preview of their potential. It is not an overstatement to say that investing in AI is a potentially trillion-dollar opportunity.¹ When Nvidia announced their Q1 earnings—$11 billion in revenue versus a $7 billion estimate—that was the iPhone moment, and there’s no turning back now.

Misconceptions and the AI value proposition

In only a few short months, the general public has become very familiar with ChatGPT. But common misconceptions remain about AI’s broader capabilities—and how it could possibly lead to trillions in earnings and GDP growth.

AI has a much wider range of applications than most people perceive. It is already proficient in language generation, image generation and recognition, reinforcement learning (feedback-based decision-making), and time series prediction (using past data to forecast future values). Where its primary value lies, however, is as a productivity booster. Across virtually every type of business, AI helps improve efficiency and increases return on investment (ROI) by enabling users to delegate time-consuming, repetitive, or unenjoyable tasks.² Potential uses include:

• Marketing: Automating processes like buying ad space; using image and scripting tools to push out ad copy across multiple platforms.
• Legal & accounting: Using language capabilities to automate document-creation and paperwork processes.
• Civil engineering: Finding efficiencies in the design and construction processes for projects like buildings and bridges; helping to intelligently carry out repetitive tasks like laying down train tracks; filling in the productivity gap as older engineers retire.
• Computing: Using copilots (AI assistants) to create documents, improve search results, or write code based on simple language prompts; Microsoft will be integrating copilot functionality directly into future versions of Windows.
• Medicine: Automatically sequencing billions of proteins, which would take humans millennia to do, in order to develop new treatments.
• Energy: Accelerating research and development on fusion technologies, which have the potential to produce virtually limitless amounts of clean energy and help solve the climate crisis.

These examples are just the tip of the iceberg. If AI continues to advance at its current rate, we predict that within two-to-three years, we’ll have what we call ‘ME TV’—a user will be able to type in a simple prompt like “create a custom rock video with me playing the guitar,” and using text-to-video capabilities, AI-generated images, and so on, the AI will be able to create a full-fledged music video. This relatively frivolous example aside, it’s not difficult to see how this type of technology could have wide-reaching applications for businesses.

AI: Boom or bubble?

It’s enticing for investors, especially those who experienced the Dot-com crash, to ask the question—is AI a real boom or just a temporary bubble? In innovation investing, which has been our area of specialization for years, we often contend with these kinds of hype cycles, which frequently pass through a period of disillusionment before eventually coming out on the other side, sometimes years later. That was the trajectory of the Dot-com cycle: valuations shot up before crashing dramatically, and investors had to wait nearly a decade before the promise of emergent technologies was realized.

In our view, the Dot-com bubble isn’t a great comparison for AI. In that instance, the crash was a function of how long it took for the commercial implementation of new technologies to take effect. There won’t be that kind of decade-long gap with AI, because the necessary infrastructure has already been built up.

As the graph below illustrates, the innovation investing cycle can be broadly divided into two periods: hype and commercialization.

Unlike Dot-coms, which spent years in the hype part of the curve, AI is already in the commercialization phase. Because of the computing power required to run AI applications, companies can’t afford to give them away to users, ChatGPT being a rare exception. As a result, they’re being immediately monetized under a subscription model, with the value far outstripping the cost—for a monthly fee of $20, for instance, a coder may be able to increase their productivity by 50%. In terms of ROI, that’s a no-brainer.

The Skynet question – A cautionary note

Remember the ‘Godfather of AI,’ Geoffrey Hinton? He recently made headlines by resigning from his position at Google and, on his way out, warning about the growing dangers of AI, stating: “Given the rate of progress, we expect things to get better quite fast. So, we need to worry about that.”³ This is a concern we hear often from investors. Is research progressing too quickly? Are we on the verge of Skynet, the world-destroying AI from the Terminator film franchise? The short answer to the Skynet question is no, but people like Hinton do raise legitimate concerns; after all, if AI can dramatically increase productivity in the workplace or lead to huge breakthroughs in the medical field, it could also be used for nefarious purposes by governments, militaries, and extra-state actors. Like all technologies, it attracts both the good and the bad.

We are already seeing some of these issues begin to materialize. Advances in AI have awoken regulators, and those debates will be ongoing. ‘Deep fakes’—realistic, AI-generated images of people and situations—are already ubiquitous in certain corners of the internet. For now, we generally assume that everything we see is real; in the near future, we’re likely to believe that everything is fake unless proven otherwise. There is also the potential for fraud, as well as the more mundane danger of inflated valuations for companies that don’t live up to their AI-related promise.

The AI industry is acutely aware of these risks. Unlike the big data era, which saw regulators react lethargically while companies flagrantly violated data privacy rules in the rush to participate in the land grab of information, the ESG experience with AI has been different. The industry has proactively created self-regulating ethics organizations to address concerns and regulators have been quicker to respond this time around. The RI team at BMO GAM is also actively working to understand and anticipate what regulatory oversight will be necessary in this new world.

The reality is that AI is a type of technology that is unprecedented in the history of humanity; we’ve never been able to replicate and automate human intelligence to this degree. Concerns like those of Hinton are well-founded, and as a society, we’ll have to work through them. As an investment opportunity, however, it is also unparalleled; the commercialization of AI is underway, and many companies are already working on authentication technology to combat deep fakes. With more monetization angles being developed and more technological developments to come, our final word for investors is this: stay tuned.

Please contact your BMO Global Asset Management wholesaler for any support and guidance.

¹ Jake Rudnitsky, “ChatGPT to Fuel $1.3 Trillion AI Market by 2032, New Report Says,” Bloomberg, June 1, 2023.

² “The economic potential of generative AI: The next productivity frontier,” McKinsey & Company, June 14, 2023.

³ Zoe Kleinman & Chris Vallance, “AI 'godfather' Geoffrey Hinton warns of dangers as he quits Google,” BBC News, May 2, 2023.