How AI Companies Can Escape the Commoditization Trap
AI Snake Oil · rss · 2026-07-10
This deep dive discusses not whether "AI is a bubble," but rather how AI companies will ultimately make money:
- Current frontier model inference businesses are prone to the "commoditization trap": model differences are shrinking, switching costs are low, prices can be freely cut, and profits are hard to sustain long-term.
- The author argues that sustainable profitability is less likely at the infrastructure layer (models/compute/data centers) and more likely by moving up the stack to application and enterprise software layers: vertical integration, deep embedding into enterprise workflows, and creating switching costs and lock-in effects.
- The article uses historical examples from capital-intensive industries like railways, electricity, telecom, cloud computing, chip manufacturing, and airlines, noting that most infrastructure providers struggle to capture the value they create.
- Conclusion: The long-term profit distribution in the AI ecosystem may depend on whether labs can achieve SaaS-like software margins rather than just selling model API calls.
More from Venture
- Moonshot AI reportedly targets a $50B round ahead of Hong Kong listing — ctjlewis · 2026-07-22
- Beth Kindig says her Q3 AI stocks report flags 5 winners up more than 100% YTD — Beth_Kindig · 2026-07-22
- AI’s economic forecasts are split by nearly a quadrillion dollars by 2035 — bittingthembits · 2026-07-22
- LLM Agent Trader pairs backtesting with model-generated trading signals — tom_doerr · 2026-07-22
- OpenRouter spend data suggests frontier models still capture most paying usage — thedealdirector · 2026-07-21
- AI is collapsing startup and operating costs as U.S. business formation hits a record — sanjaykalra · 2026-07-21