What’s Ahead: Alien Processes, Domains, and Data Models

Excerpt from the upcoming chapter on Modeling Business Processes and Domains. Reis uses the AlphaGo analogy — where Go players described playing against an “alien” — to frame how AI agents may eventually create their own processes, domains, and data models that humans cannot easily understand.

Core argument: as AI agents run thousands/millions of optimization cycles, they may develop “machine tacit knowledge” and invent processes that are statistically optimal but narratively incomprehensible. This shifts the burden to data modeling — the data model becomes the only artifact explaining what actually happened.

Reis speculates AI may invent new data formats beyond human-readable tables: high-dimensional vector spaces, dynamic ontologies modified in real-time. Mixed Model Arts becomes a hybrid discipline bridging human-defined and machine-invented models.

RDCO relevance

Forward-looking piece connecting AI agents to data modeling practice. Relevant to how we frame dbt + AI advisory work — clients will need modeling discipline more, not less, as agents enter their data workflows.