Enterprise AI is no longer a model problem. It is a data and systems problem — and incumbents are winning it.
Ensemble, in MIT Technology Review, identifies the new fault line. Model providers like OpenAI and Anthropic sell intelligence that is "general-purpose, largely stateless" and "increasingly interchangeable."1 What matters competitively is whether that intelligence resets on every prompt or accumulates over time.
Accumulation is the moat. Ensemble's model: embed the "knowledge, decisions, and reasoning" of thousands of domain experts into an AI platform. The target outcome is "higher consistency, improved throughput, and measurable operational gains" — results that "neither humans nor AI achieve independently."1 This requires operational history that AI-native startups do not have.
The architecture inverts traditional software design. An AI-native platform ingests a problem, applies accumulated domain knowledge, and executes autonomously at high confidence.1 It routes only the hardest judgment calls to human experts. The human becomes an exception handler, not the primary worker.
The "last mile" gap defines the central challenge. Most enterprise AI deployments reach a plateau of partial autonomous operation and stall. Closing that gap requires proprietary decision history — exactly what incumbents running high-volume operations have been generating for years.
A structural problem reinforces incumbent advantage. Han Xiao identifies the core LLM limitation: models hallucinate on information past their training cut-off. The fix — "forcing the model to work from verified sources" — requires verified, current operational data to exist.2 Incumbents generate it continuously.
Ensemble's counterargument to startup narratives is direct. "The prevailing narrative says nimble startups will out-innovate incumbents by building AI-native from scratch. If AI is primarily a model problem, that story holds. But in many enterprise domains, AI is a systems problem — integrations, permissions, evaluation, and change management — where advantage accrues to whomever already sits inside high-volume, high-stakes operations."1
Infrastructure investment validates the thesis. Dell's AI Data Platform with NVIDIA targets enterprise-scale data orchestration and storage for AI workloads.3 The EVOLVE26 conference circuit — Singapore, São Paulo, New York, Dubai — signals institutional capital treating AI as a permanent operational layer, not a pilot.
For enterprise buyers, the strategic question is now this: which vendor will hold the accumulated intelligence of your operations in five years.