The Economics of AI Usage and What's Next For SaaS | Benedict Evans on a16z

2026-06-08 · Source: a16z · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Software Development & Engineering · Depth: Intermediate, extended

Summary

Benedict Evans's updated "AI Eats the World" thesis, discussed on the a16z podcast, posits that foundation models are becoming commodities, with value shifting to higher application layers. He highlights agentic coding as the primary use case with strong product-market fit, leading to a significant supply crunch and pricing imbalance in AI infrastructure. Evans draws parallels to past platform shifts like mobile data, where infrastructure providers built expensive global networks but saw value accrue "up stack." He notes that while AI capex is substantial (e.g., \$700 billion from big four companies this year), it faces financial gravity limits. The discussion emphasizes that the current market disequilibrium is transitory, and the long-term profitability of model providers remains uncertain as models become more efficient and competition intensifies.

Key takeaway

For CTOs and Directors of AI/ML evaluating strategic investments, recognize that the long-term value in AI will likely reside in specialized applications built atop commoditized foundation models. Focus your development efforts on creating unique, problem-solving software that leverages AI as an underlying utility, rather than betting on the sustained differentiation or pricing power of general-purpose models. Prepare for a market where core AI infrastructure becomes a competitive necessity with diminishing returns, similar to past platform shifts.

Key insights

Foundation models are commoditizing; value shifts to specialized AI applications.

Principles

• Agentic coding demonstrates clear product-market fit.
• AI value accrues "up stack" in applications, not core models.
• Current AI pricing and supply disequilibrium is transitory.

In practice

• Prioritize building differentiated AI-native software solutions.
• Anticipate declining pricing power for raw foundation models.
• Explore AI for automating implicit, undocumented organizational tasks.

Topics

• AI Economics
• Foundation Models
• Agentic Coding
• SaaS Strategy
• Platform Shifts
• AI Infrastructure
• Commoditization

Best for: VP of Engineering/Data, Executive, AI Product Manager, Investor, CTO, Director of AI/ML

Related on AIssential

• AI Eats the World? A Reality Check with Benedict Evans — AI's rapid coding adoption highlights market disequilibrium and unresolved questions about long-term value capture.
• Making sense of the SaaSpocalypse — AI is transforming SaaS from static products to intelligent, agent-driven systems, shifting value from features to data and outcomes.
• A rational conversation on where AI is actually going | Benedict Evans — AI is a transformative platform shift, akin to the internet's early days, with value shifting to application layers and distribution.
• Europe is at risk of sleepwalking into AI armageddon — European AI startups face "armageddon" due to rising US platform costs and capital market volatility.
• $700 Billion in Capex. $50 Billion in Revenue. AI’s Math Is Broken. — Despite exponential growth, major AI labs struggle with profitability due to costs scaling faster than revenue and rapid commoditization.
• Google just fired a warning shot in the AI subscription price wars — Google's AI subscription price cut signals an accelerating commoditization of AI infrastructure, impacting pure-play providers.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by a16z.