Baron’s Ashim Mehra argues AI investors need Nvidia alongside “application” stock picks

Baron Capital portfolio manager Ashim Mehra laid out a broad view of how investors should think about artificial intelligence exposure, arguing that AI portfolios often need more than the biggest chip names to reflect how the technology is deployed across the economy.

In comments reported by Yahoo Finance, Mehra said his approach spans three categories: semiconductor leaders such as NVIDIA, smaller companies that can benefit from AI spending, and “application” stocks, meaning businesses that use AI as a feature or engine inside their own products rather than only selling infrastructure.

The framework matters because AI spending is not confined to the datacenter. Mehra’s point, as described in the report, is that markets can respond differently across the stack, so an investor seeking participation in the AI build-out may want exposure to both the suppliers of core compute and the companies that package AI capabilities for end users.

NVIDIA is the anchor in the semiconductor category in Mehra’s formulation. As the company’s brand has expanded beyond chips into an AI platform approach, the market has treated NVIDIA less like a single product company and more like a key layer in building AI systems. In that context, the idea of pairing NVIDIA with application stocks is essentially a hedge against the risk that demand or margins tighten in any one segment of the supply chain.

Mehra also suggested that smaller companies can play a role in an AI strategy, implying that parts of the ecosystem beyond the largest, most widely owned names may have additional growth opportunities. That view aligns with how investors often look for “picks and shovels” plus “picks and applications,” using different slices of the AI opportunity set to reduce reliance on a single outcome.