From Task Completer to Collaborator
Module 1 · Collaborate with AI

From Task Completer to Collaborator

Shift from using AI as a one-shot vending machine to working with it as a collaborator — and learn three practical ways to collaborate with AI on real work.

The Vending Machine Mindset

Most people start using AI the way they use a vending machine: put in a request, take whatever comes out, walk away. “Write a thank-you email.” Done.

That works for tiny tasks. But it leaves most of the value on the table, because it treats the first response as the final answer. A better model is a collaborator — a fast, tireless colleague who has read a lot but knows nothing about your situation until you tell them. With a colleague like that, you would:

  • Explain the situation before asking for help.
  • React to their first draft instead of accepting or rejecting it wholesale.
  • Ask them to push back, not just agree with you.
  • Stay responsible for what finally goes out the door.
Task-completer mindsetCollaborator mindset
One prompt, one answerA back-and-forth conversation
”Write X""Here’s my situation and goal — help me write X”
Accept or reject the outputGive feedback: “Shorter. Warmer. Drop the second point.”
AI does the thinkingAI helps you think better and faster
Output is the finished productOutput is a draft you shape and own

In practice, collaboration is a loop you go around as many times as the task needs:

Context you explain Draft AI writes Feedback you react Revise AI improves

The collaborator mindset isn’t about being polite to software. It’s about recognizing that the quality of the output depends on the quality of the conversation.


Three Ways to Collaborate with AI

Nearly every useful thing you’ll do with an AI assistant at work fits one of three roles.

1. AI as a Thought Partner

You bring a problem or a half-formed idea, and AI helps you explore it: brainstorming options, asking clarifying questions, poking holes, arguing the other side.

I'm planning a team offsite for 12 people with a $3,000 budget.
Before suggesting anything, ask me 5 questions that would help
you give better recommendations.

Asking the AI to ask you questions is one of the most underused moves there is. It surfaces the context you forgot to mention.

2. AI as a First Drafter

You know what needs to exist — an email, a policy summary, a slide outline — and AI produces a first version you then edit. The blank page is gone, and your time goes into judgment and polish instead of typing.

Draft a short email to a client letting them know their project
will be delivered one week late because a supplier missed a
deadline. Tone: apologetic but confident. Offer a 15-minute call.

3. AI as a Coach or Tutor

You want to learn or improve, not just get an output. AI explains unfamiliar concepts, gives feedback on your work, and quizzes you.

Here's the opening paragraph of my grant proposal. Don't rewrite it.
Instead, tell me the two biggest weaknesses a reviewer would notice,
and explain why.

Notice the “Don’t rewrite it” instruction. In coaching mode, you want feedback that makes you better — not a replacement that skips your learning.


Advice for Getting Started

  1. Start with low-stakes tasks you could easily check yourself — reformatting notes, drafting a routine email.
  2. Always read the output before you use it. Every time.
  3. Follow up at least once. The second response is almost always better than the first.
  4. Keep a list of prompts that worked well so you can reuse them.
  5. Be curious, not credulous. Enjoy what it can do, and stay skeptical of what it claims.
🧪 Try It: One Task, Three Roles

Open any free AI assistant (you’ll set one up properly in Module 2 — any will do for now). Pick one real task from your week, such as “prepare for a performance review conversation.” Then send three separate prompts:

  1. Thought partner: “I have a performance review next week. Ask me 5 questions that would help me prepare.”
  2. First drafter: “Draft a one-page list of my accomplishments this quarter based on these notes: [paste a few rough notes].”
  3. Coach: “Here’s how I plan to describe my biggest challenge this quarter: [one or two sentences]. Give me feedback on how it might land with my manager. Don’t rewrite it.”

Compare the three responses. Which role saved you the most time? Which one improved your thinking the most?

⚠️ Common Pitfall: Letting AI Agree With Everything

Common mistake: asking AI to review your plan and taking “This is a great plan!” as validation.

AI assistants tend to be agreeable. If you want real critique, ask for it directly: “What are the three strongest objections to this plan?” or “Argue against this as a skeptical CFO would.” A collaborator who only flatters you isn’t much help.

🧠 Knowledge Check

1. You want to get better at writing performance feedback, not just produce one piece of feedback. Which collaboration role fits best?

2. What most clearly separates a collaborator mindset from a task-completer mindset?


Up Next

Module 2 gets you hands-on: setting up a free AI workspace, understanding what the free tiers offer, and running your first structured brainstorm.