Why this matters
Most employees have used ChatGPT once or twice and concluded “it’s fine, but not life-changing.” The gap between that and “I’m meaningfully more productive at work” is small: it isn’t more clever prompts, it’s a mindset: knowing which tasks are AI-shaped, treating output as a draft, and building a habit of reps. This module sets that frame for the rest of the course.
Pick your practice model
Pick one frontier model and stick with it for the whole course. Tool-hopping kills the habit you’re trying to build.
- ChatGPT: https://chat.openai.com
- Claude: https://claude.ai
- Gemini: https://gemini.google.com
- Copilot: https://copilot.microsoft.com (use this if your employer is on Microsoft 365, it’s often pre-approved)
Quick guide:
- Most forgiving default → ChatGPT
- Heavy writing or long documents → Claude
- Google Workspace shop → Gemini
- Microsoft 365 shop with policy concerns → Copilot
The prompts in this course are tool-agnostic. They work in all four.
Which tasks are AI-shaped
Not every task benefits from AI. The wins cluster in tasks that are:
- Text-heavy: emails, docs, notes, reports, replies
- Repetitive in shape: same kind of task, different content (e.g., weekly status updates)
- First-draft work: where getting to a draft is harder than editing one
- Synthesis-heavy: pulling key points out of long material
- Translation-style: reformatting (bullets ↔ prose, casual ↔ formal, English ↔ another language)
Bad fits: tasks where AI usually loses you time:
- Anything requiring access to live internal systems AI can’t see
- Decisions that hinge on judgment, relationships, or politics
- Numerical work that needs exactness without verification (use a spreadsheet, then ask AI to explain it)
- Anything where being wrong has high cost and you can’t easily verify
Draft, don’t decide
The single most useful frame: AI drafts; you decide. You stay the editor, the approver, and the one accountable to your manager. AI gets you to a fast first draft so you can spend your judgment on what matters, the actual decision.
This frame solves three problems at once:
- Quality: you catch errors because you’re reviewing, not generating
- Trust: your manager sees your work, polished by AI, not AI’s work pretending to be yours
- Skill: you stay sharp because you’re still the decider
AI as your assistant, not your ghostwriter
Some people feel uneasy about AI helping them write. The worry usually sounds like: “Isn’t it dishonest if AI drafted it?”
Reframe it. Think of AI like a competent personal assistant who filters your inbox, drafts replies for routine messages, and surfaces the things that genuinely need your attention. That isn’t “blowing people off”, it’s the standard operating model for anyone who has ever had an executive assistant or chief of staff. You’re protecting your bandwidth so the work that requires you gets your full attention.
What separates this from ghostwriting is the editor role. You read everything before it goes out. You catch the things AI can’t know, the relationship history, the political angle, the line your manager is sensitive about. The output goes out under your name because you stand behind it, not because AI wrote it.
AI as second perspective, not oracle
A second mindset shift, harder to internalize, AI is not a crystal ball that knows the right answer. It’s a tool that cuts through noise and surfaces angles you might miss, viewpoints, perspectives, holes in your reasoning, edge cases worth considering.
Used this way, AI doesn’t replace your thinking. It pressure-tests it. You bring the question and the context; AI offers a second perspective; you decide what to keep. Over time this isn’t just faster work, it sharpens your own thinking habits. You start asking the harder questions yourself, because you’ve seen the model ask them a hundred times.
This is why “ask AI for the answer” is the wrong move. “Ask AI to challenge what I’m about to say” is the right one. We’ll come back to this frame in Modules 4 and 5.
The 70/30 rule
A reasonable target: AI gets you to 70% of a finished output. You finish the last 30%.
- 70% means structure, content blocks, first phrasing, an obvious version
- 30% means accuracy, voice, context only you have, and the call on what to keep
If you’re spending more than 30% rewriting, your prompt was too vague, go back and add context, examples, and constraints (Structured Prompt Formula from Course 1). If you’re spending less than 30%, double-check, you probably skipped verification.
Try It At Work: Pick Your Weekly Task
The fastest way to lock the habit in is to apply AI to one specific task you do every week. Reps build muscle, variety doesn’t. Pick one now and stick with it through the rest of the course.
Time: 10 min
You’ll need: Your calendar and a quick mental scan of your week.
Do this:
- List 3 tasks you do every week that involve writing, summarizing, or drafting.
- For each, ask: is this text-heavy? Repetitive in shape? First-draft work?
- Pick the one with the most checkmarks. Write it down. This is your target task for the rest of the course, and every “Try It At Work” will use it when possible.
Done when: You can name your target task in one sentence (“Every Monday I write a status update for my team.”).
Key takeaways
- AI wins on text-heavy, repetitive, first-draft, synthesis, and translation-style tasks.
- AI drafts; you decide. You stay the editor and the accountable party.
- Think of AI as a personal assistant, not a ghostwriter, you’re protecting your bandwidth, not hiding authorship.
- AI is a second perspective, not an oracle. Use it to pressure-test your thinking, not replace it.
- Aim for 70% from AI, 30% from you. More than 30% rewriting = vague prompt.
- Pick one model and one weekly target task. Reps beat variety.
Quick Check
1. Which kind of task is *least* likely to benefit from AI?
2. The "70/30 rule" suggests AI should produce roughly what share of the final output?
3. What's the main reason to pick *one* model and stick with it through this course?
4. The "AI as second perspective, not oracle" frame means