Aaron Ross

For the most direct exposure to the AI boom, maintain a core position in NVIDIA (NVDA), as it remains the "gold standard" for data center infrastructure and holds a dominant market ecosystem. Shift your focus from high-priced AI model providers like OpenAI to "Neo-Cloud" infrastructure plays like CoreWeave and Lambda, which are currently undervalued relative to their massive revenue growth. Keep a close watch for the upcoming Cerebras IPO and monitor Groq, as the market shifts from training models to the multi-billion dollar "inference" sector. Consider private or secondary market exposure to Base 10, a "hidden gem" optimizing GPU efficiency that is already generating nearly $1 billion in revenue. For a long-term "picks and shovels" play, invest in a basket of small nuclear reactor companies to capitalize on the massive electricity demands of next-generation AI data centers.

Investors should prioritize Google (GOOGL) as a strategic "hyperscaler" play, benefiting from its 14% stake in Anthropic and its proprietary TPU hardware that serves as a viable alternative to NVIDIA. For those seeking exposure to the AI energy bottleneck, focus on Small Modular Reactor (SMR) companies like X Energy or Oklo, as these fission-based technologies are closer to regulatory approval than speculative fusion. Monitor secondary markets for entry points into high-growth private firms like Anthropic, OpenAI, and Together AI, using a 34x revenue multiple as a valuation benchmark. To capture the "pick and shovel" demand of the AI supply chain, look for infrastructure providers and energy firms that have secured direct power "off-take" agreements with Amazon, Google, or Microsoft. Given the high volatility of individual AI models, the most prudent long-term strategy is to hold a diversified basket of AI-related stocks or the S&P 500 to capitalize on a multi-decade growth cycle.

Investors should prioritize the "picks and shovels" of the AI boom by targeting Semiconductors, Data Centers, and Electricity Production to capitalize on the critical global compute shortage. High-conviction private market opportunities exist in specialized GPU cloud providers like Lambda Labs, Together AI, and Crusoe, especially those with NVIDIA strategic partnerships. For exposure to rapid infrastructure execution, xAI is a top contender due to its massive Colossus GPU cluster and ability to monetize excess compute. While OpenAI faces dilution risks, Anthropic offers a more capital-efficient alternative for investors seeking pure exposure to high-quality AI models. To capture the long-term volume shift from model training to daily usage, focus on Inference Compute specialists like Groq and its high-speed LPU technology.

Investors should prioritize exposure to the AI Infrastructure sector, specifically targeting NVIDIA and upcoming IPO candidates like Lambda Labs and Together AI to capitalize on the critical global shortage of compute power. SpaceX remains a high-conviction long-term hold, with its Starlink satellite internet division projected to drive a future valuation between $2 trillion and $5 trillion as it scales to hundreds of millions of global subscribers. In the software space, Anthropic (Claude) is gaining significant enterprise momentum over OpenAI, making it a primary target for investors looking for the next dominant AI model provider. For specialized growth, Harvey AI represents a top-tier play in the legal and wealth management verticals, currently valued at $11 billion with substantial room for expansion into regulated industries. High-multiple stocks like Palantir (PLTR) and Tesla (TSLA) continue to be actionable public proxies for those seeking exposure to the aggressive "forward-valuation" growth seen in private markets for defense and AI tech.

Investors should target Anduril Industries in the pre-IPO secondary markets, as its recent $20 billion military framework and shift toward "Silicon Valley speed" position it for a potential 2.5x return or a path to a $1 trillion valuation if it expands into commercial security. For long-term exposure to "Physical AI," 1X Robotics offers a high-conviction play on humanoid robots for elder care, with a projected adoption tipping point starting in 2027. Look for Harvey AI as a dominant B2B play in the legal sector; its strategy of embedding engineers directly into client workflows creates a "moat" similar to Palantir (PLTR). In the defense sector, prioritize companies like Anduril that produce cost-effective autonomous hardware, such as the Roadrunner drone, which provides superior "attrition math" against traditional defense primes like Raytheon (RTX) and Boeing (BA). The overarching macro strategy is to invest in robotics and AI implementation firms that decouple economic growth from human labor, serving as a massive deflationary force over the next decade.