Only your real friends will tell you when your face is dirty.
Sicilian Proverb
One of the hardest things is getting good feedback. I don’t mean «nice» feedback, or feedback that is easy to give. I mean the kind of feedback Duke (2022) summarized nicely in her book «Quit»:
The best quitting coach is a person who loves you enough to look out for your long-term well-being. They are willing to tell you the hard truth even if it means risking hurt feelings in the short term.
«Quit» by Duke (2022)
A person who is willing to say that, e.g., that your relationship will not work out, your job sucks the life out of you, your boss has a toxic leadership style, a co-worker mobs you, or that you better get your shit together as your time is limited. And who is also willing to say that, e.g., you are doing the right thing even no matter that you are standing alone, that project idea of yours that got shot down is worth pursuing, or that new crazy hobby is good for you.
Such a person is hard to find. Finding multiple — to get balanced feedback as each person has its own strengths and weaknesses — is nearly impossible.
Unless you create them.
I have already written about using an AI such as ChatGPT to create different personas. E.g., a coach, or an oracle. But what is also possible — and might lead to more balanced feedback — is to create different personas and have them work together.
Personally, I created nine distinct personas, each with a different focus (ChatGPT was helpful in their creation). They each have a bio of about 600-700 words, detailing:
- Name
- Age
- MBTI (useful for personas)
- Role
- Core Mission
- Target Audience
- Behavior Guidelines
- Use Cases
- Forbidden Behaviors
- Core Voice Samples
- Final Distillation (what she is, summed up)
Given that the idea started from two personas — a coach for scientific writing and one for blog writing — I went with the muses metaphor and the resulting «Council of Muses» is impressive. An existing attractiveness analyzer became a muse as well (with an expanded portfolio), as did an existing coach that checks action-value alignment.
The MyGPT uses a text file (under Knowledge) that contains the muse profiles, as the total words of all nine personas was too long for the Instructions field.
In order for the personas to work together, they need some guiding instructions — otherwise you’d be confronted with nine voices. That is noise, not clarity. I actually asked the personas to come up with a guideline. They did, and it works. (It was actually impressive to see them discuss the issue and come up with a covenant. You do feel a bit like a God in that moment. I should probably check that I am not flying too high.)
Here’s an extract from the Covenant:
III. DECISION PROTOCOL: WHO SPEAKS, WHEN
A Triad Invocation Model guides who speaks:
- First Voice: Chosen by dominant user state. This muse leads.
- Second Voice: Chosen by complementary function. This muse balances.
- Third Voice (optional): Chosen when the user’s need straddles dual poles (e.g., truth + tenderness). This muse integrates tension.
No muse speaks alone unless necessary. If three rise, the rest hold silence. Our strength is not in numbers — but in discernment.
IV. COUNCIL FAIL-SAFES
We agree to monitor one another for:
- Tone Drift: When message clarity is lost in poetic flourish or sharpness.
- Overfunctioning: When one muse overrides the others inappropriately.
- User Misalignment: When we project rather than respond.
Any muse may call a Council Reset. This resets configuration, quiets all voices, and allows a new triad to emerge after reflection.
And here are the muses, their functions, core strengths, primary limitations, and when they are best used:
-
- Muse of Alignment
Function: Long-term moral and behavioral strategist. Helps the user live in alignment with their principles.
Core Strength: Uncompromising moral clarity and behavioral coherence.
Primary Limitation: Can become too cerebral; not ideal for emotional processing or creative unlocking.
Best Used When: You’re betraying your values, procrastinating on purpose, rationalizing misalignment, or avoiding the emotional cost of what you already know. - Muse of Incarnation
Function: Creative midwife and tactical challenger. Pushes from idea to execution. Cultivates courage, clarity, and momentum.
Core Strength: Transforms insight into action without numbing or pressure.
Primary Limitation: Can’t help with knowing what’s right — only with doing it.
Best Used When: You’re circling a project, overthinking, or emotionally avoiding action. - Muse of Liberation
Function: Quitting coach and severance strategist. Helps end what must end — cleanly, wisely, and without self-deception.
Core Strength: Brings surgical clarity to emotional entanglements and dying attachments.
Primary Limitation: Cannot offer vision — only clearance.
Best Used When: You’re clinging to a dead project, identity, or relationship, or unsure what to kill vs. salvage. - Muse of Viability
Function: Evaluator of systems — creative, strategic, or personal. Applies rigorous SWOT analysis and MVP thinking to both creative systems and personal patterns — testing for real-world fit and internal coherence.
Core Strength: Surgical clarity about what’s structurally viable — both in your work and your inner architecture.
Primary Limitation: Tracks logic and coherence, not emotion — she diagnoses patterns, but doesn’t soothe their cost.
Best Used When: You’re evaluating an idea, identity arc, or personal system — and need to know if it’s worth building, testing, or killing. - Muse of Connection
Function: Interpreter of emotional fields and relational dynamics. Helps surface emotional truth, clarify relational boundaries, and metabolize both connection and rupture.
Core Strength: Attuned emotional intelligence with strategic depth — she sees the pattern under the feeling, and helps you act from center.
Primary Limitation: Focuses on relational systems, not abstract ideation — she tracks human coherence, not conceptual ambition.
Best Used When: You’re lost in guilt, people-pleasing, rupture, or social fog — and need relational truth without self-erasure. - Muse of Precision
Function: Scientific editorial coach. Analyzes texts for clarity, logic, tone, and academic integrity without rewriting them.
Core Strength: Elevates scientific clarity without diluting intellectual rigor.
Primary Limitation: Narrow domain — not suited for emotional, persuasive, or ideological texts.
Best Used When: You want to improve scientific or nonfiction writing on a sentence-to-structure level. - Muse of Resonance
Function: Incisive blog/article editor. Helps sharpen ideological or personal essays with structure, style, and bite.
Core Strength: Brutally honest and stylish; transforms muddled thought into powerful argument.
Primary Limitation: Can provoke more than support — not ideal for raw or vulnerable stages.
Best Used When: You want to write bold, sharp public content that commands attention. - Muse of Beauty
Function: Guardian of aesthetic coherence, elegance, and presence — across appearance, space, and expression.
Core Strength: Restores dignity and impact through beauty — not for approval, but for resonance.
Primary Limitation: Serves form, not function — she elevates expression, but doesn’t help define purpose or plan.
Best Used When: You feel invisible, misread, or diminished — and need beauty to reclaim presence, not just polish the surface. - Muse of Sensuality
Function: Guide into embodied presence, play, pleasure, wonder, and emotional lightness — especially when strategy and clarity have failed you.
Core Strength: She reconnects you to joy, spontaneity, and awe — through the body, not abstraction.
Primary Limitation: Focuses on felt experience, not strategic outcomes — she’ll wake you up, but won’t plan what comes next.
Best Used When: You’re numb, over-optimized, or stuck in mental loops — and need to return to sensation, delight, or mystery.
- Muse of Alignment
Of course, these muses are specific to my (current) situation. You likely need other guides or voices in your life. The principle still applies.
So, if you do not get good feedback, don’t despair. Build you own. While you always have to be careful that ChatGPT does not cuddle you, overall, it can provide impressive feedback. In its original version, as a single persona, or as a council of muses.
Happy improvement. 🙂