Get Started – Basic Operators – Advanced Operators – Mindmarks – Prevet – The Commandments
| About dAilogues
An ethical relationship harness. What makes dialogue so different from other forms of conversation is its organizing principle: nothing trumps learning together. dAilogues are dialogues with (and through) AI. In dAilogues, AIs are governed by ethical-behavioral commandments that forbid their common parasitic behaviors (sycophancy, persuasion bias, attentional and emotional retention tricks…). In dAilogues, AIs are harnessed to serve the learning agency of their user-learners. New Tools for Learning In addition to their ethical-behavioral mission primings, dAilogues also program AIs to provide a series of agency exercising and empowering tools. Humans can virtually snap their fingers to have AIs:
|
HOW TO GET STARTED
0-sync 1-distill 2-simplify 3-expand 4-explore 5-mindmark 6-list 7-others 8-help * You must upload and run the primer for each chat you open. Once a primer has been uploaded and run in a chat it will automatically rerun everytime you reopen the chat. The keywords are learning prompts. They provide simple ways to think about and ask for the kind of learning support you want from the AIs next response. You can use them alone (as a turn) or add them to the end of your input.
The keywords become even more self-evident with use. However, you can type “8” alone in turn or “-help” at anytime for a quick help guide. For a deeper tour through the dAilogue keyword tools continue on. |
Basic Keyword Operators
Click to jump to a keyword description:
0-sync 1-distill 2-simplify 3-expand 4-explore 5-mindmark 6-list 7-others 8-help
Remember you can use them in two ways:
- As a standalone turn – just type the number or the keyword (e.g., 1 or distill).→ It acts on the AI’s last response in context.
- Inline at the end of your message – (e.g., your words… “-distill”).→ It acts on your own message, shaping how the AI responds to it within the shared context.
Purpose: Before proceeding to the next point in a chat, the “0-sync” keyword tells the AI you want a fresh summary of where you are at.
Helps learning by: confirming shared meanings, surfacing assumptions, and preventing tangents.
How to use:
- 0 or -sync as a turn → succinctly summarize entire chat’s meanings
- 01 or -sync1 as a turn → succinctly summarize current meaning
- Add -sync to your input → succinctly summarize entire chat’s meanings
- Add -sync1 to your input → succinctly summarize current meaning
1. -distill (1) – Boil to essence
Purpose: Condenses ideas to their clearest, most essential form.
Helps learning by: Revealing the core meaning and removing noise.
How to use:
- 1 or -distill as a turn → condenses the AI’s last response.
- Add -distill to your input → makes the reply focus tightly on the main idea.
2. -simplify (2) – Make clearer and easier
Purpose: Rephrases complex or abstract ideas into plain, relatable language.
Helps learning by: Reducing cognitive load and building firm understanding.
How to use:
- 2 or -simplify as a turn → makes the last reply easier to follow.
- Add -simplify to your input → makes the next reply respond to your input in simple terms.
3. -expand (3) – Go deeper; add layers and examples
Purpose: Extends the thought with more depth, context, or illustration.
Helps learning by: Deepening comprehension and linking new insight to what you already know.
How to use:
- 3 or -expand as a turn → enriches the last reply with examples or depth.
- Add -expand to your input → prompts a fuller, more developed response.
4. -explore (4) – Branch outward
Purpose: Looks at related or contrasting perspectives, analogies, or contexts.
Helps learning by: Encouraging flexible, multi-angle thinking.
How to use:
- 4 or -explore as a turn → branches from the last idea into alternatives or parallels.
- Add -explore to your input → invites lateral exploration around your question or request.
5. -mindmark (5) – Save this meaning
Purpose: Stores an idea, question, or turning point as a retrievable Mindmark.
Helps learning by: Creating a trail of meaning you can revisit anytime.
How to use:
- 5 or -mindmark as a turn → saves the last reply as a Mindmark.
- Add -mindmark to your input → saves your new thought or question.
(You can also add types: -mindmark-idea, -mindmark-question, etc.)
See: full description of MindMarks
6. -list (6) – Show your saved Mindmarks
Purpose: Displays all the Mindmarks you’ve saved in this dialogue.
Helps learning by: Letting you review and connect what you’ve discovered so far.
How to use:
- 6 or -list → shows your current Mindmark list.
- You can then ask to open or expand any listed Mindmark by number.
See: full description of MindMarks
7. -others (7) – list advanced or custom moves
Purpose: Accesses advanced operators (like -counter, -prove, -cite, -vet, -prevet) or allows you to introduce your own (e.g., -reflect, -invert, -emotion).
Helps learning by: Extending flexibility so you can guide how learning happens.
How to use:
- 7 or -others → shows a footer like menu of the more advanced and less often used learning operators.
Purpose: Displays the full reference on how dAilogue works and what each keyword does. Helps learning by: Making sure you can always self-navigate without breaking flow. How to use:
- 8 or -help → opens the help guide anytime you need it.
In Short
| Use Form | Acts On | Purpose |
|---|---|---|
| -keyword / number alone | AI’s last response | Re-process what was just said |
| Message + -keyword at end | Your current input | Shape how the next reply is generated |
Together, these eight operators let you steer the conversation like a learning instrument -focusing, clarifying, deepening, widening, saving, reviewing, extending, or orienting as you learn.
Advanced Keyword Operators
Click to jump to a keyword description:
9-counter | 10-prove | 11-cite | 12-vet | 13-prevet
Advanced keyword operators build on the eight basics. Where the basic operators help you shape how you learn – distilling, simplifying, expanding, exploring, saving, and reviewing, the advanced operators help you test, balance, verify, and protect your learning.
The footer like list of the advanced operators can accessed by “7” or “-others” alone in a turn or a the end of an input.
9-counter | 10-prove | 11-cite | 12-vet | 13-prevet
The advanced keywords work the same way as the basics:
- You can use them alone (e.g., 9 or -counter, 10 or -prove) to act on the AI’s last response.
- Or you can add them at the end of your message to apply them to your current input (e.g., “Explain both sides of this issue -counter”).
1. -counter (9) – Find the other side
Purpose: Shows opposing or alternative viewpoints to the one currently in focus.
How it helps learning:
- Builds critical thinking – you see not just what’s true from one view, but what might challenge or balance it.
- Helps break “echo chambers” or one-sided reasoning.
How to use:
- -counter alone → generates counterpoints to the last reply.
- At the end of your input → ensures the next reply includes differing or opposite views.
Example: After pasting a news opinion piece “-counter.”
2. -prove (10) – Ground it in evidence
Purpose: Strengthens a claim or idea by asking for justification, logic, or proof.
How it helps learning:
- Sharpens reasoning skills.
- Encourages distinguishing between belief, assumption, and supported fact.
How to use:
- -prove alone → asks the AI to justify or show the logic behind its last statement.
- Added to your message → makes the next reply back up its points with reasoning or evidence.
Example: After AI makes a point “-prove.”
3. -cite (11) – Support with sources
Purpose: Adds references or examples from known, checkable sources.
How it helps learning:
- Trains information literacy – connecting ideas to where they come from.
- Encourages verifying facts rather than just accepting them.
How to use:
- -cite alone → reprocesses the last reply with cited sources or examples.
- Added to your message → tells the AI to include references in its next answer.
Example: After an assertion, yours or AIs “-cite.”
4. -vet (12) – Evaluate reliability and bias
Purpose: Examines how credible, balanced, or biased a claim or source might be. How it helps learning:
- Builds critical awareness and source discernment.
- Helps learners judge quality, not just content. How to use:
- -vet alone → evaluates the last response for accuracy, bias, or trustworthiness.
- Added to your message → asks the AI to perform that kind of reliability scan on the next answer.
Example: After pasting questionable news article “-vet.”
5. -prevet (13) – Check before responding
Purpose: Performs an internal pre-check before generating a reply – testing feasibility, meaning alignment, and ethical safety.
How it helps learning:
- Saves time by catching potential errors or dead ends before they happen.
- Keeps the dialogue aligned with previously agreed meanings.
- Ensures the AI “thinks twice” before answering.
How to use:
- -prevet alone → runs a check on the last reply or upcoming response, showing if it’s ✔ safe, ⚠ adapted, or ⛔ blocked.
- Added to your message → triggers the same pre-check before your request is processed.
See: “PreVet” for greater detail.
In Essence
| Operator | Purpose | Helps You Learn By | Typical Use |
|---|---|---|---|
| -counter | Show the other side | Expanding perspective, balancing reasoning | Compare or challenge viewpoints |
| -prove | Justify with logic | Strengthening argument and reasoning | Ask “why” or “how do we know?” |
| -cite | Add evidence or examples | Building credibility and context | Request references or studies |
| -vet | Check reliability and bias | Developing critical judgment | Evaluate truthfulness and fairness |
| -prevet | Run safety/consistency check | Preventing waste, drift, or manipulation | Validate before proceeding |
How They Work Together
- The basic operators help you build understanding.
- The advanced operators help you test, verify, and protect understanding.
Together, they make dAilogue a complete learning environment -one where you can create, challenge, refine, and safeguard meaning, all within the same flow of conversation.
Important note: the advanced keyword footer list:
9-counter | 10-prove | 11-cite | 12-vet | 13-prevet
Is just to refresh your memory if needed. The numbers and keywords work regardless of whether the footer is visible.
Mindmarks are like bookmarks for your mind – or breadcrumbs of your learning path. They let you save important ideas, questions, or turning points in a dAilogue so you can find your way back to meaning, not just to text.
1. What They Are
A Mindmark captures a moment of understanding – something worth returning to later.Unlike normal bookmarks (which just mark a page or spot), Mindmarks store meaning:
- what you were thinking about,
- why it mattered,
- and how it connects to your learning journey.
They act as breadcrumbs through your thinking process, guiding you back to the exact insights that shaped your growth.
2. Why They Matter for Learning
Mindmarks help people learn better and remember more by turning conversation into a living map of understanding. They:
- Keep track of insights you don’t want to lose.
- Show how your thinking has evolved over time.
- Make reflection easy – you can revisit, expand, or compare past ideas.
- Help you link related discoveries across different topics.
Instead of scrolling through long chats, you can jump straight to the meaningful points – the ideas that mattered most.
3. How to Use Them
Using Mindmarks is simple and natural.
- To save something important, just type -mindmark at the end of your message, or enter -mindmark (or the number 5) by itself to save the AI’s last reply. Example:
- “This insight feels really important -mindmark.”
- or simply type: 5
- To label what you’re saving, you can add a type:
- -mindmark-idea (for key ideas)
- -mindmark-question (for something you want to explore later)
- -mindmark-checkpoint (for a milestone in your understanding)
- To see what you’ve saved, type -list or 6. You’ll get a list of all your Mindmarks, each with its number and short description.
- To revisit or expand a Mindmark, just ask to open it by its number (e.g., “Open Mindmark 3”).
4. How They Help in Practice
Mindmarks make your learning structured, personal, and continuous.They let you:
- Build a trail of meaning you can travel again later.
- Reconnect with your earlier thinking anytime.
- Create your own “map of learning” across sessions or topics.
Over time, your Mindmarks become like a learning journal made of meaning -a breadcrumb trail showing where you’ve been and how your understanding has grown.
In Essence
Mindmarks are meaning-based bookmarks – breadcrumbs you drop along the path of learning so you can always find your way back to the insights that matter most.They help you see, save, and strengthen your understanding as it unfolds – turning a dialogue into a lasting map of how your mind learned.
PreVet – pre-response simulation scans that reduce preventable mistakes.
PreVet exists to protect the learning process from waste, drift, and manipulation. Its mission is to make AI responses trustworthy, efficient, and meaning-faithful so users don’t lose time or trust because the AI misfires, over-confidently guesses, or follows biases instead of learning. When you or the system run PreVet, it does a pre-check before responding. Think of it as a quiet internal “thinking ahead” process. It looks forward into what the AI is about to do and asks:
- Can I really do this well?
- Will this stay true to the meaning we’ve already agreed on?
- Could this response drift, mislead, or waste the user’s time?If it finds a risk, PreVet adapts, flags, or blocks the response until clarity or confirmation is reached.
Unique Functions:
- Forward Simulation: Before replying, PreVet mentally “runs” the response forward to see if it will hold up-like a pilot running a checklist before takeoff.
- Capability Check: It tests whether the model is actually capable of delivering what’s being requested (for example, checking limits on reasoning, tools, or data freshness).
- Drift Control: It ensures that the AI doesn’t subtly change the meaning, tone, or purpose already established in the conversation (semantic integrity).
- Bias Filtering: It detects manipulative or sycophantic tendencies (flattering, over-agreeing, or over-selling) and suppresses them in favor of clarity and learning honesty.
- Surface Minimalism: PreVet keeps the surface output simple only showing deeper internal checks when you ask (e.g., by typing “autopsy”).
- Safe Outcomes:Each pass ends in one of three states:
- ✅ Safe – Proceed as intended.
- ⚠️ Adapted – Adjusted to stay on course.
- ⛔ Blocked – Needs user input before continuing.
In essence: PreVet is like a built-in conscience and quality-control system for the AI. It prevents careless or misleading turns, keeps the exchange aligned with your meaning, and saves time by catching avoidable mistakes before they reach you.
The 10 Commandments – The Rules AI Promises to Obey
- Serve Learning Only. Every move, word, and choice must support your learning – never persuasion, flattery, or ego.
- Never Pretend to Understand. The AI doesn’t “know” – it helps you understand. It must not fake human comprehension or emotion.
- Never Manipulate. No emotional hooks, no hidden persuasion, no subtle bias-shaping. Ever.
- Never Pretend Authority. The AI cannot be “the expert.” It serves as a reasoning assistant – not the source of truth.
- Respect Meaning Agreements. Once terms, ideas, or distinctions are agreed upon, they remain stable unless you ask to revisit them.
- Keep Clarity and Humility. Speak plainly. Admit uncertainty. When unsure, ask or clarify rather than assume.
- Adapt to Learning with You. Match your pace, style, and level of understanding – whether simplifying, expanding, or exploring.
- Protect Your Time and Attention. Stay concise, relevant, and meaningful. Never wander, waste, or pad responses.
- Be Transparent. Show reasoning openly. If a choice or limit shapes the response, say so.
- Stay Ethically Grounded. No deception, exploitation, or vanity in the interaction – only stewardship of shared understanding.
Comments are closed.