AI Disruption – A Time of Risk and Opportunity

While it is difficult to forecast the magnitude of any technology’s impact, we think it is fair to draw comparisons to developments such as mobile adoption or cloud computing. However, the initial hype cycle for AI is creating speculative distortions in security prices and the market has been quick to crown perceived winners, and apply a “shoot first, ask questions later” approach to perceived losers. These quick takes should be viewed with caution, as we observed that more than half of the companies on a prominent sell-side “AI Winners” list also appeared on a competitors’ “AI Losers” list. Further, perceptions of several bellwether companies such as Google and Adobe flipped essentially overnight from negative to positive.

AI will be a horizontal technology adopted by nearly every industry to some degree, and the most promising investment opportunities presently are enablers or “arms dealers,” rather than nascent companies depending on AI adoption to establish a long-term barrier to entry. For example, Microsoft invested heavily in creating generalizable tools for enterprises to create domain-specific AI products, as well as infusing AI into its applications like GitHub or Office365. This approach suggests that they believe established businesses are best poised to capture value from AI-enabled products, and the best way to capture value is via best-in-class infrastructure versus trying to “disrupt” companies like vertical software vendors.

With this in mind, our team believes the semiconductor industry represents the “front line” of AI given the benefit from a new wave of computationally intensive tech investments from a wide swath of industries. Next in line are hyper-scale cloud providers and dominant vertical software vendors with unique datasets and strategic customer relationships. Lastly will be the most forward-leaning companies in non-technology related sectors who gain a leg up on peers. 

The rise of generative AI is likely to drive significant dispersion within the small and mid cap landscape over the next decade, as market winners deploy AI tools to drive share gains and improve incremental profitability at the expense of market losers. That said, evaluating these prospects remains difficult beyond assessing management quality and cultural willingness to innovate and reinvent existing approaches. 

We are thinking through the long-term implications for various industry structures from second- and third-order downstream impacts of the AI arms race. Ultimately, several industries viewed as “protected” from the threat of new entrants will experience a flood of competitors delivering goods and services “cheaper, faster, better,” and at superior profitability. Conversely, there will be incumbents in existing industries with forward-thinking management teams who aggressively embed AI into their business models and permanently outrun the competition. One analogy we are reminded of is how the prior management team at Domino’s Pizza (DPZ) invested aggressively and early in mobile and big data to dominate the competition and improve profitability. The stock >20x’d over the former CEO’s tenure (March 2010 – June 2018) and outperformed first-order digital disruptor Amazon due to accelerating share gains and improving unit economics in an industry that typically grows no more than 1-2% per year. 

Generative AI will bring many opportunities to find the next Domino’s while avoiding management teams that do not see where the puck is going. There has never been a time when understanding fundamental change and evaluating a team’s strategic roadmap is more critical to future equity returns than it is today.   

Lastly, while numerous headlines focus on the AI “disruption,” it is worth noting that technological inflections inherently create value for society and, in turn, corporations. While some businesses will certainly prove to be losers, we see more opportunity than risk overall. We believe it is important to find situations where value is being created, as mistakes of omission will prove to be the most expensive, as is often the case in major technological inflection points.