How to play, and not play, the AI boom

An important lesson the Kaufmann team has learned in all our years of investing in innovation is transformative breakthroughs rarely are investable right away. Think of the cryptocurrency wave that only just crested—decades after the earliest conceptions of digital currency. If you have Delta miles or Marriott points, you are still reaping the benefits of the breakthrough that many deemed a “fad.” Or back further to the late ’90s and early aughts, when the early internet (remember the World Wide Web?) eventually gave way to the dot.com boom … and bust. Today’s venture capital market rose from the ashes of that bust. Shotgun sequencing took decades of grudging work before the human genome project moved to the next level—and today, that project is already helping save lives. All proved to be great outcomes for society and our economy. But for asset managers, they presented a lot of stones in the road, with some diamonds at the end.

Almost by design, innovations are disrupters in their fields, and it takes time to separate the wheat from the chaff—the investments that are going to turn into profitable opportunities from the ones that fall by the wayside. Companies are built to survive, not always to thrive, and very few can maintain their growth throughout multiple business and innovation cycles—just ask Motorola, PalmPilot, JDS Uniphase, Myspace and Dell. Consider even Google Search, which still has ads supplementing its revenue model to enable a seemingly simple function, as well as Gmail, which you may be surprised to learn is not profitable despite its dominant market position. Often, it’s the technologies that survive while the companies churn or fail.

Identifying these transformative breakthroughs in advance is no easy feat—if it were, everyone would have bought Bitcoin in 2010 (when to get out is a different question). But there are signs we’re in the early innings of the next big tech breakthrough: generative artificial intelligence (AI). At its core, this represents the next iteration of the cloud computing revolution that emerged from the dot.com bubble, which grew out of the mainframe revolution, which evolved from the microchip—I think you get the point. It falls under the broad umbrella of “machine learning.” ChatGPT, for example, does its computing automatically without human coders or enablers. With its sheer volume of potential applications and uses, AI has captured the imagination of investors, consumers and governments the world over.

Indeed, markets are practically euphoric over the many “what ifs” related to generative AI and its potential for productivity gains, the creation of entirely new industries, etc. The list seems endless. But as with anything new in the economy, much fear on many fronts has accompanied this breakthrough. Just think of the movie, “I, Robot.” There are many others. The reality is the amount of automation behind AI’s rise, and its potential to build on itself, makes the unknown a powerful fear factor. “Disruption” may be exciting for the disrupters, not so much for the disrupted.

Who stands to benefit the most from generative AI?

Our take is the big companies with the most data and computing power at their disposal will be first in line to capitalize on AI’s potential. However, the pace of change is likely to be rapid. Who the major winners might be in the long term remains uncertain. So far, hardware enablers look like immediate beneficiaries. Just consider the sheer amount of infrastructure that will be needed for the massive data center buildout that’s surely coming. Hardware enablers also need electricity from multiple sources and complicated transformers to enable the input, creating spinoff beneficiaries in these fields. And so on ….

One thing to keep in mind is generative AI users have little preference for “legacy” brands. Sheer ability is, and likely will continue to be, the deciding factor. We’ve seen a number of newcomers rise from nowhere and gain more traction than some of the incumbent firms that have been deemed “AI winners” by sell-side analysts and the financial media. And let’s not forget, the bottom line is about, well, the bottom line. History is rife with seemingly amazing innovators and innovations that proved unsustainable.

Specialization and uses potentially could be very narrow

One early area of focus we’re watching is biotech, where some of AI technologies will enable drugs to be tested in a supercomputer before even going to phase 1 trials. This could cut costs by half by some estimates. The potential applications are practically endless. Think of what generative AI could do for companies that assist diabetics. It could possibly anticipate eating habits and respond to fluctuations in blood sugar patterns with appropriate levels of insulin without a human’s input at all.

The bottom line (in the metaphorical sense this time): for all its potential, generative AI is unlikely to have a significant near-term impact on the economy at large. Why? This innovation is clearly deflationary. And what have rates done since this all started: They’ve gone up, not down. All this hype could turn into deflation, which could cause the market to rethink its rally. That said, only time will tell if generative AI becomes analogous to the mainframe computer age led by IBM, the search and email age with Yahoo or even the social networking boom kickstarted by Myspace. Let’s all stay tuned.