A common claim is that the flood of frontier intelligence will drown the application layer. There will be no moats left to build. That framing is incomplete.

Moats are useless against a flood. What matters are dams and waterways: systems that direct intelligence toward crops, reservoirs, and homes that would not otherwise receive it. Intelligence is a utility with unbounded demand. The world does not want raw models. It wants problems resolved and outcomes achieved.

Over the next five years, every company will use AI either to run critical operations or as part of the product it sells. In both cases, generic intelligence will not be enough.

Off-the-shelf models are useful because they are general. But the deeper AI is embedded into a business, the more company-specific details matter: policies, workflows, risk tolerance, escalation rules, historical patterns, and customer context. These specifics do not live in a generic model’s weights.

In the current AI landscape, many builders and investors feel despair because frontier models seem poised to absorb almost everything. But according to Sarah Guo, this view misses a critical category: untrainable work.

Measurable, public, and legible tasks (coding benchmarks, generic customer support scripts, standardized processes) are being commoditized rapidly from both below (open models) and above (labs folding scaffolding into weights). What remains valuable and defensible is work that is private, context-rich, high-stakes, and dependent on trust, permissions, accountability, and deep integration with proprietary systems.

The labs are aggressively pursuing the obvious horizontal layer of agentic AI — general code generation, writing, image creation, and simple tool-calling agents. This is the well-trodden Yellow Brick Road, where raw model capability dominates and labs hold decisive advantages in scale, distribution, and architecture.

The real moats and durable value creation lie in The Rest of Oz: complex, vertical, multi-step workflows involving messy real-world data, legacy systems, regulatory constraints, multi-player coordination, and industry-specific tribal knowledge. These require deep scaffolding, data flywheels, governance, and deterministic outcomes that horizontal models cannot easily replicate.

Enterprise SaaS companies are reporting that AI is not disrupting their core businesses — it is dramatically expanding them.

Q1 2026 earnings calls revealed five consistent themes:

AI can fake effort. It cannot fake time.

Software features and intelligence can now be cloned in days. The enduring advantages are those accumulated slowly through real-world operation and physical reality.