Advances in Artificial Intelligence (AI) have ushered in a new era of innovation in technology. For investors, AI has become a hard-to-ignore market theme. Given the size and scale of market movements and the breathless coverage in the media, it is no wonder that everyone from the retail investor to sophisticated institutional investors has a point of view.

Our investment team believes the risk in trying to predict the leaders in AI is not currently justified by potential for meaningful returns. Instead, we believe investment opportunities exist in budding AI beneficiaries, who do not have the same all-or-nothing risk attached to them. From our perspective, these beneficiaries include both companies in the AI supply chain who stand to benefit regardless of which platforms become leaders as well as companies who stand to benefit from applying new AI technologies to improve their business growth or profitability.

What Is Artificial Intelligence (AI)?

We define traditional AI simply as automation often coupled with machine learning to perform specific tasks. That automation can take many forms. Think of smartphone applications like Google Maps, or language translators, or autonomous cars. The machine is trained to analyze data and offer an optimized solution, like the fastest route home given the traffic of the day. This type of predictive AI has already been shown to be effective in improving efficiency and continues to be widely adopted.

What Is Generative AI?

Generative AI is less well-developed and requires significantly more data to become an effective tool. Most generative AI models today are trained on large quantities of data and when given a prompt by the user generate new data. Models are making word-by-word predictions using the context from a user’s prompt and all the data they were trained on. The promise of generative AI is that with specific inputs, you can receive high quality analysis and content.

Current State of AI Investing & Challenges of AI

AI investing is at an early stage, with companies currently focused on building out the needed infrastructure. An acronym commonly used to describe the phases of technological transitions is “IPA”, or Infrastructure, Platform, and Application. We are currently in the infrastructure stage.

In our view, this early phase has benefitted larger companies in the data center, hardware, and semiconductor sectors. Because AI requires such large amounts of computing power and data to train the models as well as engineering talent to design them, only the largest and most-capitalized companies have been able to pursue training the most expensive AI models, what are often referred to as “Foundational Models.” Foundational Models are called this because they seek to provide the foundational intelligence that future AI applications will later be built upon, and presently only large companies can afford to train those Foundational Models that will have the greatest impact on generative AI going forward.

While the infrastructure layer of AI is being built out, enterprise companies are setting aside capital to invest in the new technology. The challenge is we still do not have a unified Platform (e.g. Microsoft Windows) or many Applications (e.g. Microsoft Word). Companies are still experimenting with how to apply AI.

We believe it could take time to see use cases that will justify the tremendous spending we are seeing today. Investors and companies are waiting for the emergence of the “killer application” that could lead to new businesses that would generate material revenue. While we understand the promise of AI is to make us more productive or efficient, how that technology is monetized at scale remains uncertain. In our view, we see three truly viable enterprise use cases for generative AI today: chat bots, summarization, and automation. While these have the potential to provide cost savings for businesses adopting them, we have not yet seen the true potential of generative AI’s impact on enterprises.

We Believe KAR Companies Have Unique Data Assets to Leverage AI

Despite this uncertainty regarding generative AI’s applications, timeline, and the viability of returns, we do have confidence that proprietary data will be a differentiated competitive advantage as organizations seek to train generative AI models while keeping their business intelligence protected and secure. Companies that own unique proprietary data that cannot be replicated or accessed by competitors will retain that differentiation regardless of how the AI technology landscape ultimately plays out.

We will continue to actively monitor how AI changes the competitive landscape as well as the opportunities it presents for new investments. That said, we do not have a mandate to invest in AI technology specifically. Kayne Anderson Rudnick’s primary focus is researching and seeking out the highest-quality businesses that can adapt in any type of market to generate meaningful returns through an entire cycle. This has been our approach since 1984 and has produced high conviction portfolios through several technology transitions in the past.

Contact Kayne Anderson Rudnick today to speak with our team about our investment strategy and download our Thinking About Artificial Intelligence paper to learn more about how Kayne Anderson Rudnick examines the growth and influence of AI and provide our thoughts on how to invest in this evolving technology.