AI isn’t like other tech cycles — and investors may need a new approach

AI may dominate headlines, but we believe people are still underestimating its potential for disruptive growth and attractive investment returns. At the same time, investors risk falling into familiar traps — hoping broad index exposure will capture the opportunities on offer and struggling to distinguish the disruptors from the disrupted. 

While the potential pitfalls remain the same, AI is unlike any previous technology paradigm shift in several ways:

• Hardware that can access AI applications and services is nearly ubiquitous globally
• AI is advancing at an unprecedented pace relative to past technology paradigm shifts
• AI technology is much more scalable than other technological innovations
• Companies are investing in AI on a massive scale

The reason this matters is that AI-related disruption could present ample opportunities for active managers to add value through investments in technology and other sectors — but selectivity is key. Underestimating the pace and scale of the disruption could lead to missed opportunities for investors. Let’s dig into each of these differentiating factors in more detail. 

What differentiates AI-induced disruption from prior cycles?

1) The ubiquity of hardware makes the adoption of software today much faster than in the past

Past technology transitions required large hardware cycles with costs borne by enterprises or individuals. This time, consumers already have the devices and connectivity necessary to adopt the technology immediately on its release. Two-thirds of the global population have internet access, with many countries having nearly 100% access. Enterprises have access to cloud-service providers that can scale large deployments at a rapid pace.

As a result, we expect this technology to be adopted at a far faster pace than prior technology cycles. Consider the fact that it only took OpenAI’s ChatGPT two months to reach 100 million monthly users after launching in December 2022. 

2) The rapid pace of technological advancements, and subsequent cost savings

Since the launch of ChatGPT, we have been stunned by how quickly large language models are improving. Critically, reasoning models generate their own training data as each thought a model works through can then be fed back into the model, effectively learning in real time as it solves the problem.

In addition to this, the cost of technology is falling rapidly due to advancements in semiconductors, driving even more profitable use cases for generative AI. Just a few years ago, reasoning models would have been prohibitively expensive due to their compute-intensive nature. 

3) AI technology scales more easily than other technological advances 

If we think back to previous technology cycles, both the automation of labor and the augmentation of knowledge work – meaning tasks that involve thinking, analyzing or creating – via computers required new equipment and processes. Automating assembly lines has been happening for over a century, moving slowly because the ability to reduce costs in a physical world takes time. Augmenting knowledge work via computers also takes equipment and skills (learning to type or use spreadsheets, for example) and while this still happened rapidly, it took decades. 

With AI, we are seeing the augmentation of work grow at an unprecedented rate. As we move toward an agentic world (the first use cases of which went live this year), we are seeing a technology that can replace humans at scale with less friction than any prior automation cycle. This is happening fast, and is likely to accelerate, in our view.

4) Large technology companies are investing at an unparalleled pace 

Large technology companies and AI labs are investing at a historically unparalleled pace to build this technology. This race to scale models and win business is accelerating the pace of development and adoption of this technology.

Why this matters

Few are aware of the extent of the disruption that will occur over the coming years and, in our view, policymakers and investors alike are not prepared for the shift at hand. For example, we expect that a significant portion of entry-level white-collar jobs will be automated over the next few years. As reasoning models get better, particularly at data analytics, we believe the number of impacted roles will grow significantly. It is important to note that this initial disruption is happening today even before the widespread adoption of AI agents, which will entirely replace large segments of the workforce.

In our view, this is not being accurately captured in the mental frameworks or modeling currently being applied to investing. We are optimistic that eventually there will be substantial job creation, as we saw with the adoption of spreadsheets, but that is unlikely to happen contemporaneously in large enough effect to offset losses. 

What this could mean for investors

The disruption sparked by AI will offer opportunities for investors, but we see three potential pitfalls: 

1) Not all technology companies will be winners

Some of the largest technology companies will likely emerge as winners, but newer companies will also succeed. It’s possible the company that comes to define success in the AI era hasn’t even been founded yet. In our view, deep sector expertise is needed to discern the true winners and losers.

2) Capital spending doesn’t necessarily equal returns

The fact that tech companies are investing on a massive scale doesn’t necessarily mean they’re guaranteed to succeed. Investors should closely analyze which firms are spending wisely and which ones aren’t to discern whether they are using capital effectively. 

3) US exceptionalism concerns don’t mean that US companies are any less exceptional — but it’s wise to look elsewhere too

A handful of large technology stocks have dominated the US equity market for years. Almost none of them were top performers at the height of the dot-com era in the early 2000s. We don’t yet know how different the list of best-performing companies will look in the future, as AI capabilities and integration scale up. 

This isn’t a death knell for today’s US-based “Magnificent Seven,” but it is a reminder to guard against complacency. Exceptional US companies may continue to excel, but in this nascent stage of AI, there’s plenty of room for new competitors — in the US and elsewhere — to grow, succeed, and even displace incumbents. Prudent investors will look for opportunities among new and established companies alike, both in historically dominant markets, like the US, and in others, like Asia, where innovative, new AI-focused companies have long runways to thrive. 

Taken together, these pitfalls and the opportunities investors may be able to access by avoiding them are a strong reminder of the potential value in an active approach. Compared to their passive peers that buy indiscriminately based on an index, active managers with the power to scrutinize company fundamentals in search of those with the brightest futures may be better positioned to capitalize on all this unprecedented tech cycle has to offer.