Even with capital and data, incumbents struggle to compete with focused AI startups because of cultural inertia. Existing go-to-market strategies, sales compensation, org structures, and obligations to a large customer base are fundamental laws of physics that prevent large companies from moving at startup speed.
Previous computing abstractions provided resources like storage or compute, but the human programmer retained control over the core logic. AI represents a fundamental shift where users abdicate reasoning and logic to a non-deterministic, third-party system, asking it for the answer without defining the precise steps or even the exact desired end state.
Previously, startups competed on agility while incumbents held capital and distribution advantages. In the AI era, startups with massive funding can directly challenge incumbents on a capital basis. This, combined with AI solving distribution and the incumbent's cultural inertia, creates a new competitive dynamic.
While AI solving long-standing math problems is impressive, its real value is questionable if those problems lack economic significance. The key indicator of AI's impact is its ability to solve problems that unlock tangible economic value, a test many current 'breakthroughs' have yet to pass.
While the mechanics of AI models are understood (they are in-distribution pattern matchers), we have no precedent for predicting the emergent capabilities of a single digital artifact built with billions of dollars of compute. The conversation must shift from how they work to what these unprecedentedly scaled artifacts can actually do.
The notion that VC is a zero-sum game with too much capital chasing too few deals is flawed. More available capital allows companies to stay private longer, accruing more value on the private side. Furthermore, new technology waves like AI can productively absorb massive capital infusions, effectively growing the total addressable market for venture investment.
Counterintuitively, mathematicians are among the most excited by AI's progress in their field. They view AI not as a replacement but as a powerful new level of abstraction, similar to the invention of calculus or calculators. It automates tedious work, allowing them to explore new frontiers of thought and discovery.
For decades, tech innovation was engineering-bound, as hiring more engineers didn't linearly increase output (the 'Mythical Man-Month'). AI flips this paradigm. A small team can now productively deploy massive amounts of capital on compute, shifting the primary constraint from engineering talent to capital availability.
The history of computing has cycled through different constraints. It began as capital-bound (acquiring expensive mainframes), shifted to engineering-bound with the rise of software, and has now returned to being capital-bound due to the massive compute costs for training AI models. This historical context reframes today's landscape.