The Information’s AI stack matrix answers the crucial question: where does value actually concentrate across the AI stack? The answer surprises those expecting simple “picks and shovels” or “application layer” investment theses.
The matrix maps value capture across infrastructure, models, middleware, and applications. What emerges isn’t a single dominant layer but a complex pattern of valuedistribution that shifts as the stack matures.
Infrastructure captured early value—NVIDIA’s dominance. But infrastructure advantages commoditize as alternatives emerge and hyperscalers bring capabilities in-house. The commoditization cycle is already visible in inference pricing.
Foundation models captured the next wave—OpenAI, Anthropic valuations. But model capabilities are converging, and open-source alternatives narrow the gap. Model differentiation may prove temporary.
Middleware and applications represent the emerging opportunity. As lower layers commoditize, value migrates up the stack. The companies that capture durable value will likely be those solving specific business problems rather than general AI capabilities.
The Investment Implication
Stack analysis suggests a barbell approach: position in infrastructure during dominance windows (NVIDIA today), but prepare to shift as commoditization arrives. Long-term value likely concentrates at the application layer where network effects and customer relationships create defensibility that technical advantages cannot.
The picks-and-shovels thesis was right for 2023-2024. The application thesis may be right for 2025-2026.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
