At the Frontier: Insight on AI and Technology Stocks

Like the technology itself, the investment story in artificial intelligence (AI) is continuously evolving ― presenting exciting opportunity for active investors like Tony Kim, head of the Fundamental Equities Global Technology team.

He and 30 BlackRock colleagues traveled 300 miles over five days in June to meet with leaders of 25 public and private technology companies in San Francisco and Silicon Valley. The annual “tech tour” ― the 12th of its kind ― is a critical input to the team’s fundamental research process. Mr. Kim recently joined The Bid podcast, along with Michael Gates, lead portfolio manager of BlackRock's Target Allocation Models, to reflect on the marathon week and a pace of innovation that has AI quickly pushing the edges of its vast potential.

Another year, unprecedented progress

When it comes to AI, one year can amount to decades of progress. While the first nine years of the tech tour had few whispers of AI, the past three have featured an increasingly loud AI takeover. Even relative to 2024, when Mr. Kim said the conversations centered on ideas like chatbots and AI agents, this year brought attention to “the edge of what is possible.”

“Super intelligence, AI at the frontier, was the prevailing narrative,” he recalls, noting that this represents a move up the AI “stack.” This stack begins with infrastructure at the foundation, comprising the chips and cloud framework that enable AI. This is followed by the intelligence layer, which includes the data and AI models, and caps at the application layer, the software services and solutions that leverage the AI intelligence to perform work.

Mr. Kim and his team, who view companies and investment opportunities across the full AI stack, are now seeing the extension of AI applications into physical, real-world forms.

To be sure, the rates of change in AI are “unprecedented” and potentially hard for some to imagine. As humans, he says, we're accustomed to linear forms of change in increments of 5%-10%. In AI, he sees change at the chip level coming in at 2x a year, and at 10x a year at the model intelligence layer. In all, it’s a 20x potential improvement with an exponential scaling of capability. “You compound that one year, two years, three years, four years, and you are soon in the thousands of times improvement in just a few years.”*