Resources
Glossary
Plain-language definitions for key terms used across the Phronesis Intelligence Series, the framework, and FridayLocalAI.
This glossary helps readers understand how terms such as artificial intelligence, prompt, model, agent, memory, artifact, source discipline, governed intelligence, local-first AI, and human authority are used throughout the series.
Purpose
Clear terms support responsible AI use.
Artificial intelligence is surrounded by fast-moving language. The same words can mean different things depending on whether the conversation is technical, commercial, academic, creative, or practical.
The Phronesis Intelligence framework uses terms in a specific way: to keep AI-assisted work connected to practical wisdom, human authority, source discipline, governance, and local-first control.
This glossary is not intended to settle every technical debate. It gives readers a working vocabulary for understanding the books, resources, and FridayLocalAI material.
The goal is clarity before complexity.
Core AI Terms
Artificial intelligence and related terms
These terms describe the tools, systems, and working behaviors readers encounter when using AI.
Artificial Intelligence
A broad term for computer systems that perform tasks associated with human intelligence, such as generating language, recognizing patterns, classifying information, summarizing content, or supporting decisions.
AI Assistant
A user-facing system that helps with tasks such as drafting, summarizing, explaining, planning, coding, analysis, brainstorming, or document work through conversational interaction.
Model
The underlying trained system that produces responses, classifications, summaries, predictions, or generated content. In ordinary use, the model is one part of the larger tool experience.
Large Language Model
A model trained to process and generate language. Large language models can produce fluent text, but fluency does not guarantee accuracy, current information, source support, or judgment.
Prompt
The instruction, question, context, or task given to an AI system. A prompt can be simple, but serious work often requires purpose, audience, source material, constraints, and review standards.
Response
The output produced by an AI assistant. A response may be useful as a draft, explanation, suggestion, or starting point, but it still requires human review before serious use.
Agent
An AI-supported system or workflow that may perform tasks with a defined role, scope, tool access, memory, or instructions. Agents require governance because expanded capability can create expanded risk.
Workflow
A repeatable process for using AI in a structured way, such as drafting, reviewing, checking sources, revising, preserving an artifact, and making a final human decision.
Framework Terms
Phronesis Intelligence concepts
These terms describe the practical-wisdom framework used across the series.
Phronesis
Practical wisdom: judgment applied in real situations. In this framework, phronesis asks what should be done, under whose authority, with what evidence, for what purpose, and with what regard for consequence.
Phronesis Intelligence
The framework and publication identity focused on practical wisdom in AI-assisted work. It emphasizes human authority, source discipline, governed intelligence, and local-first control.
Practical Wisdom
The disciplined ability to decide what should be done, why it matters, what evidence supports it, what risks exist, and whether the work should be used.
Human Authority
The principle that people remain responsible for accepting, revising, rejecting, publishing, preserving, implementing, or relying on AI-assisted work.
Source Discipline
The practice of identifying claims, checking evidence, reviewing citations, distinguishing assumptions from facts, and verifying information before using AI-assisted output.
Governed Intelligence
AI-assisted work placed inside visible boundaries of authority, memory, prompts, artifacts, sources, access, traceability, and human approval.
Local-First AI
An approach that begins with local control over records, memory, artifacts, source context, and authority before external tools or services are connected.
Final Judgment
The human decision that determines whether AI-assisted output should be accepted, revised, verified, escalated, preserved, published, implemented, or rejected.
Governance and Records
Terms for durable AI-assisted work
These terms describe how AI output becomes work that can be found, reviewed, corrected, and carried forward.
Artifact
A work product created, shaped, or preserved during AI-assisted work, such as a document, report, checklist, summary, table, prompt, decision record, or generated file.
Memory
Stored context that may help future work. Memory can be useful, but it must be scoped, reviewable, correctable, and distinguished from approved authority.
Record
Preserved information that helps explain what happened, what was decided, what source material mattered, what changed, and what should guide future work.
Traceability
The ability to understand how work developed, including prompt context, source material, generated output, review decisions, artifact versions, and final use.
Recoverable Decision
A decision preserved clearly enough that future work can understand what was decided, why it was decided, what scope applies, and whether it remains active.
Approved Work
AI-assisted output that has been reviewed and accepted by the person or organization with authority for a defined purpose.
Superseded Material
A prior draft, instruction, memory, decision, or artifact that has been replaced and should no longer guide current work unless explicitly referenced for history.
System of Record
The place where authoritative records live. For local-first AI, the system of record should remain under governed operator control whenever the work matters.
Practice Terms
Terms used in AI review and method
These terms support practical exercises and reader workflows.
Context Frame
A structured description of the task, audience, purpose, source material, constraints, desired format, and review standard before AI generates output.
Evidence Pass
A review that identifies claims, classifies risk, checks source needs, verifies currency, separates experience from evidence, and removes unsupported statements.
Red-Team Review
A deliberate challenge pass that looks for weaknesses, missing assumptions, risks, unsupported claims, unclear logic, or ways the work could fail.
Privacy Review
A check that asks what material is sensitive, what should be redacted, what should remain local, and what should not be shared with external tools.
Authorship Review
A review that asks what AI assisted with, what the human changed, what sources deserve credit, whether disclosure is needed, and whether the final work can be defended.
Final Judgment Review
A final human review that determines whether output is ready for use, needs verification, should be revised, should be preserved, or should be rejected.
FridayLocalAI Terms
Platform-related language
FridayLocalAI is described as a governed local-first AI workbench because the platform is designed around work, not only answers.
FridayLocalAI connects conversations, prompts, memory, source context, artifacts, records, models, tools, and human approval into a more durable working environment.
These terms help readers understand how the platform relates to the Phronesis Intelligence framework and the third book, Practical Wisdom in the Machine.
Product-specific details and updates belong on FridayLocalAI.com.
Visit FridayLocalAI.comRelated terms
- Workbench
- Local-first control
- Governed memory
- Artifact handling
- Source context
- Human approval
Continue Exploring
Use the glossary with the books and resources.
The glossary supports the book series, framework pages, FridayLocalAI material, reader guides, and AI practice tools.
For practical exercises, continue to AI Practice Tools.
For the larger structure of the series, continue to Books or Framework.
For platform-specific information, continue to FridayLocalAI.com.
Clear language strengthens practical wisdom.
Use this glossary as a working vocabulary for responsible AI-assisted work, source discipline, governed intelligence, and human authority.