Key Concept • 2019
Unknown Knowns
Issues on the edge of human and machine understanding that remain unresolved in smart technology—the boundary where AI capabilities and human cognition diverge.
Status
Used Extensively
Year Referenced
2019
Domain
AI & Machine Learning
Knowledge Graph
Understanding Unknown Knowns
The concept of “Unknown Knowns” originates from the Rumsfeld matrix—a framework distinguishing what we know we know, what we know we don’t know, what we don’t know we don’t know, and critically, what we don’t know we know. In the context of smart technology, Joseph Byrum applies this framework to identify the boundary conditions where AI systems and human cognition diverge in fundamental ways.
Unknown Knowns in AI represent capabilities and knowledge that exist within systems but remain inaccessible or unrecognized. For humans, this includes intuitive knowledge we possess but cannot articulate. For machines, it encompasses patterns detected in data that lack interpretable meaning. This gap creates dangerous blind spots in automated decision-making systems where neither human operators nor AI systems fully understand what the other “knows.”
Understanding these boundaries is essential for building effective human-AI collaboration. Byrum’s Understanding Smart Technology series dedicates three installments specifically to Unknown Knowns, examining how human minds and machine minds process information differently, and identifying the dangers that emerge when these cognitive gaps go unaddressed in automated systems.
Related Articles
From the Understanding Smart Technology series exploring Unknown Knowns
INFORMS Analytics
The Human Mind Unknown Knowns: Understanding Smart Technology Part 5
Exploring the intuitive knowledge humans possess but cannot articulate to machines.
INFORMS Analytics
The Machine Mind Unknown Knowns: Understanding Smart Technology Part 6
How AI systems detect patterns without interpretable understanding of their meaning.
INFORMS Analytics
Smart Machine Dangers Unknown Knowns: Understanding Smart Technology Part 7
The risks that emerge when human-machine cognitive gaps go unaddressed.
Related Course
Understanding Smart Technology
9-part series exploring AI capabilities and limitations
Frequently Asked Questions
What are Unknown Knowns in the context of AI?
Unknown Knowns represent the boundary between human and machine cognition—knowledge that exists but isn’t accessible or recognized. For AI systems, this includes patterns detected in data that lack interpretable meaning. For humans, it encompasses intuitive knowledge we possess but cannot articulate to machines. This creates dangerous blind spots in automated decision-making.
Where does the term Unknown Knowns come from?
The term originates from the Rumsfeld matrix, a framework distinguishing four categories of knowledge: known knowns, known unknowns, unknown unknowns, and unknown knowns. While Donald Rumsfeld famously discussed the first three categories in 2002, philosopher Slavoj Žižek highlighted “unknown knowns” as the neglected fourth quadrant—knowledge we possess but don’t realize we have.
Why are Unknown Knowns dangerous in smart technology?
Unknown Knowns create dangerous blind spots because neither human operators nor AI systems fully understand what the other “knows.” An AI might detect patterns that operators can’t interpret, while humans possess tacit knowledge they can’t encode into systems. When these gaps go unrecognized, automated systems can fail in unexpected ways, particularly in novel situations where this hidden knowledge would be critical.
How does this concept relate to the Intelligent Enterprise?
The Intelligent Enterprise framework emphasizes AI augmentation over replacement precisely because of Unknown Knowns. By keeping humans in the loop through the Iron Man Model approach, organizations can leverage both machine pattern recognition and human intuition. This addresses the Unknown Knowns problem by creating systems where human and machine cognition complement rather than substitute for each other.
External References
Explore Joseph Byrum’s complete body of work on AI strategy and organizational transformation.