What the Machine Inherits

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

What the Machine Inherits Why the Future of Artificial Intelligence Is Also a Question About Us Much of the current conversation surrounding artificial intelligence is organized around a single anxiety: what happens when AI begins to act without us? It is an important question. Contemporary AI systems are already becoming more agentic. They can plan, use tools, execute multi-step tasks, preserve context, coordinate actions, and—increasingly—operate for meaningful periods without a human approving every intermediate decision. In robotics, those capabilities are beginning to cross from the symbolic world into physical space. But I have found myself returning to a different question. What does artificial intelligence inherit before it begins to act? That question changes the frame. Artificial intelligence does not arrive from outside human civilization. It is trained within our languages, institutions, histories, preferences, examples, reward structures, economic systems, cultural assumptions, and disagreements. Even as AI becomes increasingly autonomous in operation, it remains the product of an informational ecology that is overwhelmingly human in origin. The machine may increasingly determine how to execute an action. But before that autonomy emerges, something has already shaped what the system notices, what it interprets as success, what it learns to avoid, what it remembers, what it is permitted to do, and what kinds of behavior are reinforced. That is the foundation of what I have begun calling normative inheritance.