We scan new podcasts and send you the top 5 insights daily.
Lawson thinks foundation-model companies have grown faster and absorbed more application value than most observers expected three or four years ago. He can count on one hand the scaled AI application companies like Harvey, Sierra, and coding tools. The models' improving reasoning keeps unlocking new categories, but fewer app winners have emerged than predicted.
Data reveals an extreme power law where model labs OpenAI and Anthropic capture nearly all AI startup revenue, and their share is growing. This indicates value is accruing to the foundational layer, posing an existential threat to the long-term viability of application-focused startups.
Foundational AI models will commoditize into a utility layer where companies buy "intelligence on the fly." The real, sustainable profit will be captured by application companies that leverage various models to solve specific business problems, as most enterprises lack the expertise to use raw models effectively.
As base AI models become more capable and multi-modal, they pose an existential threat to application-layer startups. Users can now execute complex workflows with a single prompt that previously required a stack of specialized tools, making many applications redundant.
Similar to how blockchain protocols like Bitcoin and Ethereum accrued more value than the apps built on them, AI foundation models are getting 'fatter.' They are absorbing more capabilities, allowing users to perform complex tasks in a single step within the base model, reducing the need for specialized application-layer companies.
Counter to fears that foundation models will obsolete all apps, AI startups can build defensible businesses by embedding AI into unique workflows, owning the customer relationship, and creating network effects. This mirrors how top App Store apps succeeded despite Apple's platform dominance.
The belief that a single, god-level foundation model would dominate has proven false. Horowitz points to successful AI applications like Cursor, which uses 13 different models. This shows that value lies in the complex orchestration and design at the application layer, not just in having the largest single model.
Startups building on top of AI models, like coding assistant Cursor, are extremely vulnerable. As foundation model companies like Anthropic improve their own native capabilities (e.g., Claude Code), they can quickly capture the market and render specialized tools obsolete.
Unlike software bottlenecked by engineering headcount, AI models scale with capital. A frontier model company can raise more than its entire app ecosystem combined, then use that capital to launch competitive first-party apps and subsume third-party developers.
The battleground for AI startups is constantly shrinking like the map in Fortnite. Foundation models like Anthropic's Claude are aggressively absorbing features, turning what was a standalone product into a native capability overnight. This creates extreme existential risk for application-layer companies.
The economic value in AI is rapidly shifting away from foundational models, which are becoming commoditized far faster than anticipated. The real, sustainable business models are emerging at the infrastructure layer (cloud, chips) and the application layer, not in the foundational models themselves.