🧪 AI Lab after session Notes + Templates + Recording (OPEN AI LAB SESSION #1)

Hey you! Lucky you… you got a 🎁 Goodies Bag!

Thanks for joining the Open AI Lab Session #1!

After every experiment in the Lab, our guests (aka Test Subjects #∞) leave with something special:

🎁 The Goodies Bag - your post-lab surprise kit with all the essentials to keep building and exploring on your own.

You will also find the session recording below.

Let’s get inside today’s bag:

🧠 OPEN AI LAB SESSION #1 NOTES:

Welcome to the AI LAB. These Notes are your end‑to‑end companion: they connect every concept, show you exactly what to click, and give you examples you can copy‑paste.

What you’ll get from this guide 

- A clear story spine for using AI:
Discover → Organize → Prompt → Build → Automate/Agent → Research → Decide → Act.

- Step‑by‑step instructions with screenshots cues (what to click, where to find things).

- Copy‑ready prompts, templates, and checklists.

- A complete cross‑platform comparison so you know which tool to use for what.

- The Los Menchos Squad: my specialized AI clones and how to use them.

Keep this open while you follow along. Skim the bold lines first, then dive into the parts you need.

1) ⏯️ REPLAY THE SESSION!

2) 🚨 DOWNLOAD DECK USED DURING THE SESSION:

3) AI LAB SESSION #1 NOTES:

0) The Story Spine — How everything connects

1) Discover opportunities (Fruit Picker GPT) →

2) Organize your workspace (Project Folders) →

3) Prompt like a pro (templates + voice) →

4) Build your own assistant (Custom GPT) →

5) Automate with tasks & agent workflows →

6) Research deeply with citations →

7) Decide & Act (pick 1 small win today; plan next 7 days).

Outcome: less time on boring tasks, more time on creative/strategic work.

1) Quick Concepts (to sound smart and be smart)

AI vs ML vs LLM

AI = the umbrella (machines doing intelligent things).

ML = how AI learns from data.

 LLM = a kind of ML that talks (ChatGPT/Claude/Gemini style).

> Think: AI = Brain · ML = Learner · LLM = Talker

Agents vs Agentic Workflows

· Agent = does one specific task (e.g., “Summarize this PDF”).
· Agentic = plans & executes multi‑step work (e.g., “Draft a launch plan, create slides, schedule emails”).
> You’ll see this when we use Projects, Tasks, and tool access together in ChatGPT.

Key players (who to reach for and when)

· OpenAI (ChatGPT) — generalist, great UX, Custom GPTs, Projects, Tasks.

· Anthropic (Claude) — thoughtful reasoning, long documents, safety.

· Google (Gemini) — strong Workspace integration, web/search grounding, images.

· Meta (Llama) — open‑source models powering many startups/tools.

· Microsoft (Copilot) — deep Microsoft 365 integration across Outlook/Word/Teams.

Reality check: Choose the tool that fits the job. Many pros mix two: e.g., Claude for big docs + ChatGPT for workflow/creation.

Cross‑Platform Comparison (high‑level)

Symbols: ✅ strong · ➖ partial/limited · (varies by plan). Feature names change fast; focus on capability, not labels.

2) Your First 60‑Minute Playbook

10 min — Discover: Run Mencho AI #15 — The Fruit Picker (link below) and capture 3 quick wins.
10 min — Organize: Create AI Project: [Your Topic] with a Control Panel chat and 2 working chats.
15 min — Prompt: Use the Role + Goal + Task + Context + Format template to produce a real deliverable.
15 min — Build: Draft a simple Custom GPT to repeat that deliverable reliably.
10 min — Automate: Turn the deliverable into Tasks, and (if available) wire simple agent steps.

3) DISCOVER — Mencho AI #15 “The Fruit Picker” 🍊

What it does: Interviews you about your role, routines, pains → returns a personalized AI Opportunity Map with low‑effort, high‑impact wins.

Use it:
1. Open the Fruit Picker link.
2. Answer the questions honestly (work + life).
3. Copy the Opportunity Map into your notes.

Pro tip: Pick one idea you can do today in 15 minutes.

This GPT is part of the Los Menchos Squad (see below).

4) ORGANIZE — Project Folders (ChatGPT)

Goal: make your AI a reusable workspace, not a series of one‑off chats.

Setup (2 minutes): 
- Create folder: AI Project: [Topic].
- Create chat → rename Control Panel (📌 pin it).
- Add 2–3 working chats (e.g., Content Strategy, Presentation Builder, Inbox Drafts).
- Paste links (Notion, Docs, Sheets) at the top of relevant chats.

Control Panel content (copy):
- Objective & constraints.
- Voice/tone guide.
- Key audiences.
- Reusable prompts/macros.

Why it works: Continuity + speed + consistency → your second brain.

5) PROMPT — Talk so AI delivers

Universal Template:

Act as a [ROLE] and help me [GOAL] by [TASK].
Here’s context you need: [CONTEXT / INPUTS].
Format it like this: [OUTPUT STYLE/STRUCTURE].
Ask clarifying questions before answering.

Voice Mode option: just say it.
Example:
> “I’m an ops manager. Take this messy to‑do list and convert it into a two‑column plan: Urgent vs Important, with estimates.”

Three copy‑paste examples: 

- Ops/Procurement:
Act as a Lean ops coach… prioritize this backlog; create a risk watchlist; propose 3 automations.
- Comms/Marketing:
Act as a comms strategist… turn these notes into a Q3 content calendar; draft 2 emails; 6 social posts.
- Product/PM/Education:
Act as a product thinker… map assumptions; create a one‑page strategy; outline a stakeholder deck.

6) BUILD — Your first Custom GPT (repeatable results)

Why: When you find a prompt that works, “bottle it” so anyone (or future‑you) can reuse it.

Steps: 
1) Explore GPTs → Create.
2) Name & Purpose (e.g., Newsletter Draft Buddy).
3) Instructions: paste a refined version of your best prompt + guardrails.
4) Starter message: a friendly hello + what it needs from user.
5) Knowledge (optional): upload examples, tone guide, FAQs.
6) Tools (optional): enable browsing, code, or connectors as needed.
7) Save & test with 2 very different inputs.

Quality checks: 
- Does it ask clarifying questions?
- Are outputs consistent?
- Can a teammate use it without you?

7) AUTOMATE — Tasks & Agent workflows (ChatGPT focus)

A) Tasks (lightweight automation)

What: Convert any response into a to‑do list you can track.

How: 
1) Ask for a plan (e.g., “Draft my 4‑week onboarding schedule”).
2) Convert to Tasks (UI: plus/…”Create tasks”).
3) Edit, mark done, or regenerate.

Habit ideas: 
- Nightly check‑in: “At 9pm, ask if I trained, stretched, took creatine.”
- Weekly planning: “Every Sunday, propose a calendar based on these priorities.”

Note: If reminders/scheduling aren’t available in your plan, create calendar events from the task list, or use a Zap/automation.

B) Agentic workflows (when available)

Idea: Chain steps so the model can plan → draft → refine → package.

In ChatGPT Projects (when enabled): 
- Keep files + context in the project.
- Ask: “Create a plan → then generate the slide outline → then draft the email.”
- If your plan includes tool access (web, code, connectors), the model can fetch, transform, and assemble.

Showcase: - “Plan my event → create a timeline → generate task list → draft invites.”

8) RESEARCH — Deep Research (how it works & how to steer it)

What it is: A mode where ChatGPT breaks your question into sub‑questions, searches the web, reads sources, and returns a sourced report.

Turn it on: - If you see Deep Research or Browse/Search in your model/tools menu, enable it for the chat. (Availability varies by plan.)

Run a query: 
> “Deep‑research the impact of AI on disability support services. I need: 5 key trends, risks, and 3 practical pilots for a mid‑size org. Cite sources.”

What the model does under the hood: 
1) Plans sub‑questions.
2) Searches & opens likely sources.
3) Extracts/compares claims.
4) Drafts a structured answer.
5) Attaches citations and links.

How to steer quality: 
- Constrain scope (“AU/EU only”, “last 18 months”).
- Ask for structure (bullets, table, exec summary).
- Request contrasts (“industry vs academia views”).
- Spot‑check 1–2 sources before using.

Great use cases: 
- Build presentation backgrounds (with sources).
- Compare vendors/tools.
- Create FAQ briefs for leadership.

9) The Los Menchos Squad — Specialized Clones

I use a team of specialized GPT clones. A few you’ll meet:

10) Your 7‑Day + 90‑Day Plan

Next 7 days 
- Pick 1 Fruit Picker win and do it in 15–30 min.
- Set up your AI Project folder.
- Save 1 Custom GPT (even if simple).
- Turn your deliverable into Tasks.

Next 90 days 
- Month 1: stabilize 3 recurring AI wins (content, inbox, planning).
- Month 2: add agentic chains (plan → draft → package).
- Month 3: standardize (templates, SOPs, team training).

🚨 Limited-Time Offer: AI Lab 40% OFF

🔥 You were in the room. You get the deal.

Use this code to join the full AI Lab at 40% off:
SECRETCODE (expires in 48h)

GET YOUR SPOT

We go deep, build real stuff, and it’s designed for non-technical humans who want to stay relevant, fast.

That’s it, test subject.

Now go experiment. 🧪
And if you liked this... bring a friend next time.

Mencho
🪬 

What do you think of this edition? Tap your choice below and let me hear it 👇

🪬🪬🪬🪬🪬 AMAZING - LOVED IT!

🪬🪬🪬 GOOD / OKAY

🪬 NOT GREAT

OR… HIT REPLY and share your thoughts on what you want to have more of in this Newsletter!

FOLLOW ME ON LINKEDIN I POST ABOUT AI!