Framework
The Phronesis Intelligence Framework
Phronesis Intelligence is a practical-wisdom framework for using, reviewing, preserving, and governing artificial intelligence in serious human work.
The framework asks a question that capability alone cannot answer: when AI produces something useful, what makes that output responsible, supported, recoverable, and still under human authority?
What It Means
Practical wisdom applied to intelligent systems
Phronesis means practical wisdom: judgment applied in real situations.
In the context of artificial intelligence, practical wisdom asks what should be done, why it should be done, who has authority, what evidence supports the work, what consequences may follow, and what should be preserved or rejected before the output becomes part of real work.
Phronesis Intelligence does not treat AI as an oracle, a person, or a replacement for human judgment. It treats AI as a powerful working instrument that must operate inside human responsibility, source discipline, ethical boundaries, and recoverable systems.
The goal is not simply to generate more output. The goal is to help people turn AI-assisted output into work that can be questioned, corrected, sourced, preserved, and responsibly used.
Framework Principles
Five principles for responsible AI-assisted work
The framework organizes AI use around human responsibility rather than machine confidence.
Practical Wisdom
Ask not only what AI can produce, but whether the output is useful, responsible, appropriate, supported, and worth using.
Human Authority
AI may assist, suggest, summarize, draft, and review. People remain responsible for what is accepted, published, implemented, or relied upon.
Source Discipline
Fluent output is not proof. Serious work requires source checking, claim review, uncertainty, and clear separation between fact and assumption.
Governed Intelligence
Memory, prompts, artifacts, agents, models, tools, and decisions need visible boundaries so the system does not confuse suggestion with authority.
Local-First Control
When ownership, privacy, continuity, and recovery matter, the location of data, memory, records, and runtime authority matters too.
Human Authority
The person remains responsible for the work.
The framework begins with a simple boundary: AI can assist judgment, but it cannot replace responsibility.
A generated answer may be clear, polished, organized, and persuasive. That does not make it true, complete, current, ethical, private enough to use, properly sourced, or appropriate for the audience.
Human authority means the person or organization using AI remains responsible for deciding what should be trusted, revised, removed, escalated, preserved, published, implemented, or rejected.
That authority should not be hidden in the background. It should be visible in the workflow: review passes, source checks, privacy decisions, approval steps, preserved records, and final human judgment.
Explore Human AuthorityThe machine can help.
It can draft.
It can summarize.
It can compare.
It can suggest.
But the human must decide.
Source Discipline
A confident answer still needs evidence.
AI systems can produce polished language faster than people can verify it. That makes source discipline central to responsible use.
The framework asks users to separate facts, assumptions, interpretations, recommendations, and unsupported claims. It encourages readers to ask what evidence is needed, which claims require current verification, and whether an output should be revised or removed before use.
This is especially important when work involves current information, public claims, legal or financial issues, medical topics, academic use, technical implementation, publication, or business decisions.
Explore Source DisciplineA source-aware question set
- What claims are being made?
- Which claims need evidence?
- Which facts may have changed?
- What source would actually support this statement?
- What is assumption, interpretation, or recommendation?
- What should not be used until verified?
Governed Intelligence
Memory, prompts, artifacts, and decisions need structure.
A serious AI workflow cannot depend only on a long chat thread.
As AI-assisted work grows, conversations begin to carry decisions, drafts, source material, project direction, and records that future work may depend on. Without governance, useful fragments can become confusing, outdated, or falsely authoritative.
Governed intelligence means placing AI output inside structures that preserve meaning: projects, records, summaries, artifacts, source maps, memory notes, prompt layers, review steps, and clear authority boundaries.
The framework does not ask users to overcomplicate every task. It asks users to match the level of governance to the seriousness of the work.
Explore Governed IntelligenceGovernance distinguishes:
- Draft from decision
- Suggestion from authority
- Memory from record
- Source from generated language
- Exploration from execution
- Convenience from control
Local-First AI
Where the work lives affects who governs it.
Local-first AI is often discussed as a privacy issue. In the Phronesis Intelligence framework, it is also an authority issue.
When AI-assisted work becomes part of long-term projects, business planning, intellectual property, documentation, memory, source repositories, or institutional knowledge, the location of that work matters. So do backup, recovery, inspection, access, ownership, and runtime control.
Local-first does not mean rejecting every external service. It means establishing a governed foundation first, then connecting outward deliberately under human authority.
Explore Local-First AILocal-first asks:
- Where does the work live?
- Who controls access?
- How is memory preserved?
- How are records recovered?
- What depends on external platforms?
- What remains under the operator’s authority?
How the Books Use the Framework
The framework develops across the series
Each book approaches Phronesis Intelligence from a different angle: lived collaboration, practical method, and governed system design.
Book 1
What We Learned Together
The reflective foundation. This book shows how sustained AI collaboration revealed the need for memory, continuity, source discipline, human judgment, and governed intelligence.
Book 2
After the Prompt
The practical method book. This volume teaches readers how to manage the work after an AI response appears: context, review, evidence, privacy, authorship, and final judgment.
Book 3
Practical Wisdom in the Machine
The architecture and governance volume. This book connects Phronesis Intelligence to FridayLocalAI, local-first systems, memory, artifacts, authority, and durable AI-assisted work.
FridayLocalAI
The framework becomes practical in systems.
FridayLocalAI is the local-first workbench being developed from the same principles.
The framework explains why serious AI work needs human authority, memory governance, artifacts, prompt discipline, source awareness, and recoverable records. FridayLocalAI is the platform effort that brings those concerns into a working environment.
PhronesisIntelligence.com explains the books and framework. FridayLocalAI.com remains the product and platform destination.
Learn About FridayLocalAIFramework to platform
Framework: practical wisdom, source discipline, human authority, governed memory.
Platform: local-first workbench, artifacts, records, prompts, models, and operational continuity.
Purpose: make AI-assisted work more durable, inspectable, and accountable.
Practical wisdom is the discipline around capability.
The Phronesis Intelligence framework exists to keep AI-assisted work useful, responsible, source-aware, recoverable, and under human authority.