Filomena no és una becària – Wind Cat Training

Philomena She's not an intern: she's your collaborator 🤝

Objective: how to relate to AI as a collaborator, working elbow to elbow: with role, limits and feedback.

The anecdote (naming it… and re-naming it)

The first day I asked him: “What do you mean?” It came out Yellow: “because it gives light and helps to see”.

We were doing great until he showed up. Grok to the feed and… hateful comparisons.

We opted for Philomena: clearer, more ours. Tell him/her “Filomena, act as editor and help me polish this for non-tech teams" It makes me think of her as a companion, not as a guessing machine.

When you have a role, a goal and standards, things flow.

And none beef with Grok: he has rockets; Filomena, notebook and candy canes. 😉

The alliance with AI (key principles)

Role: "acts as [editor/teacher/trainer]".

Short brief: objective, audience, context, format.

Iterate: outline → draft → polish.

Feedback +/Δ: + what works; Δ What changes?

Verification & limits: do you want data contrast or just style? and what it can NOT do.

What does it mean? +/Δ (fast and useful feedback)
“+” = Keep / Like / Works.
“Δ” (delta) = Change / Improve / Adjust.

Mini-template (copy/paste)
Feedback +/Δ
+ [3 things that work]
Δ [2–3 specific changes: “shorten…”, “add CTA…”, “change tone to…”]

Express example (LinkedIn post):
+ Clear title and appropriate tone.
Δ Reduce to 180–220 characters and add a final question.

Express recipe: ELBOW

Context · Orobjective · Llindars (extension, “good enough”) · Zstyle wave (tone/voice) · Esimple

Template (copy/paste)
Context: teams 40+, little tech; LinkedIn.

Objective: post with 4 tips, publishable.

Thresholds: 180–220 words, final question, no jargon.

Style zone: close, humor 7/10, playful 9/10.

Examples: I like the tone of this [referent].

Act as expert in XXSS. Pound 1st version + 3 titles.

The most common human errors (and how to avoid them)

Treat it like Google (1 blank line) → ELBOW in 5 lines.

Wanting perfection at 1st → ask scheme first.

No tone example → contributes 1–2 references.

Fuzzy feedback (“I don’t like it”) → +/Δ concrete.

Do not verify data → indicate if you want contrast.

Briefing that changes mid-game → “new version: change of audience to X”.

Technical tips that help (inspired by Jon Hernández's talks)

Think in steps: first skeleton, then fill, at the end polish.

Context that weighs: attach key points or a mini-dossier (3–5 docs) so that the AI has a basis.

“Branded Philomena”: create custom GPTs/instructions with style guide and fixed examples for repetitive tasks.

Automate what is repeated: when the flow works (brief → draft → review → publication), connect it with tools like Make/Zapier.

Credits: these approaches to repeatable process, good briefing and pragmatic automation are ideas that Jon Hernandez often defend; we are inspired to adapt them to our daily lives.

Cycle 3× (fast)

Outline → Draft → Polish (closes with CTA and data check).

PS: CTA means Call To Action (call to action). It is a short sentence what does it say to the person What do you want me to do now?.

Gemma Mateu Vergely · Wind Cat Training

💬 Let's talk!

And you, what? role would you ask Filomena today? Leave me your ELBOW brief and we turn it into a publishable result.

Tell me in the comments on Linkedin https://www.linkedin.com/in/gemma-mateu-vergely/

or Instagram https://www.instagram.com/giapermortals/

And if you want to continue the series “Me and the GIA”, give me follow and turn on notifications.

#IA #GIA #Productivity #Learning #XSocial Networks #Humor #LinkedInCatalà #FilomenaGIA #GIAperMortals #WindCatTraining #IAAmbSentit