What is the core concept of "fluent AI use"?	Holding a small, clear mental model of the machine to recognize and respond to the type of failure when something goes wrong.
Name the four properties that define AI's capabilities and limitations.	Token Prediction, Knowledge, Working Memory, and Steerability.
According to the framework, what is the primary limitation regarding AI's "Knowledge"?	Difficulty with rare, post-cutoff, niche, local, or contested topics.
What type of instruction maximizes "Steerability" and user control?	Short, concrete, and verifiable instructions (e.g., "respond as a table").
What is the primary limitation of AI's "Working Memory"?	Handling very long documents or expecting cross-session continuity.
How are the Capabilities & Limitations Framework and the 4D Framework related?	They are two sides of one system; the properties explain what the 4D competencies are responding to.
What does "calibrated trust" require a user to do?	Locate their task on each continuum and match their verification and context habits to where it sits.
What is the ultimate goal of synthesizing the four properties and training fingerprints?	To build a working mental model of the AI's operational boundaries.
When assessing a task, what is the recommended immediate action after learning this framework?	Identifying one concrete change or mitigation step to implement in the AI practice this week.
What does the framework suggest about the nature of the properties themselves?	While the exact boundaries will shift, the shape of these four properties remains useful for understanding AI behavior.