---
title: "Your First Week with AI: A Practical Guide"
description: "Day-by-day plan for your first week using AI tools. No theory. Just what to do, in order, with specific outcomes."
pillar: "AI Fundamentals"
level: "beginner"
date: "2026-01-20"
url: "https://theglitch.ai/academy/fundamentals/first-week-with-ai"
---

# Your First Week with AI: A Practical Guide

Day-by-day plan for your first week using AI tools. No theory. Just what to do, in order, with specific outcomes.


# Your First Week with AI: A Practical Guide

> **The Glitch's Take:** "Most people spend months 'learning AI.' You need seven days. Deploy, don't study."

**Part of:** [AI Fundamentals: The No-Bullshit Beginner's Guide](/articles/fundamentals/ai-fundamentals-guide)
**Level:** Beginner
**Reading Time:** 10 minutes

---

## The Point

You don't need to understand how neural networks work. You don't need to read research papers. You need to start using AI tools on real tasks, today.

This is a seven-day plan. Each day has a specific task, a tool to use, and a concrete outcome. By day seven, you'll have practical AI skills—not theoretical knowledge you'll forget.

---

## Before You Start

**Minimum setup:**
- Claude.ai free account (sign up at claude.ai)
- 30 minutes per day

**Optional but recommended:**
- Claude Pro subscription ($20/month) — removes rate limits
- Perplexity free account — for research days

That's it. Don't install anything else yet.

---

## Day 1: Your First Real Task

**Goal:** Complete one actual work task with AI assistance.

**Time:** 30-45 minutes

### What to Do

1. Pick a task you'd normally do yourself:
   - Write an email
   - Summarize a document
   - Outline a presentation
   - Draft meeting notes

2. Open Claude.ai

3. Describe the task clearly:
```
I need to write an email to [recipient] about [topic].
Context: [relevant background]
Tone: [professional/casual/formal]
Length: [short/medium/detailed]
```

4. Review the output. Edit it. Use it.

### Today's Outcome

One completed task. That's it. Not perfect—completed.

### What You'll Learn

- How to give clear instructions
- Where AI output needs human editing
- The actual time savings (or not)

---

## Day 2: Iteration and Refinement

**Goal:** Learn to improve AI outputs through conversation.

**Time:** 30 minutes

### What to Do

1. Start a new task in Claude (something different from Day 1)

2. After the first response, give feedback:
```
This is good, but:
- Make it more concise
- Add specific examples
- Change the tone to be more direct
```

3. Continue the conversation. Ask for alternatives:
```
Give me three different versions of the opening paragraph.
```

4. Keep refining until you're satisfied.

### Today's Outcome

One polished piece of content through 3-5 rounds of iteration.

### Key Insight

AI gets better with feedback. The first output is a draft, not the answer.

---

## Day 3: Research Mode

**Goal:** Use AI to research a topic you care about.

**Time:** 45 minutes

### What to Do

1. Open Perplexity (or Claude with web access)

2. Research something relevant to your work:
```
What are the current best practices for [your field topic]?
Focus on changes in the last 6 months.
```

3. Follow up with specific questions:
```
What are the main criticisms of this approach?
Who are the leading practitioners?
What tools do they recommend?
```

4. Ask for sources:
```
List the sources you're drawing from.
```

### Today's Outcome

A research brief on one topic with verified sources.

### What You'll Notice

- AI research is faster than manual Googling
- But you need to verify claims, especially for recent information
- Follow-up questions unlock depth

---

## Day 4: Templates and Systems

**Goal:** Create a reusable template for a recurring task.

**Time:** 45 minutes

### What to Do

1. Identify something you do repeatedly:
   - Weekly status updates
   - Meeting summaries
   - Client emails
   - Content outlines

2. Create a template prompt in Claude:
```
I'm going to give you [input type]. Your job is to [output description].

Format:
[Specify exact format you want]

Style:
[Specify tone, length, structure]

Example of good output:
[Paste an example]
```

3. Test it with real input

4. Save the prompt somewhere you can reuse it (notes app, document)

### Today's Outcome

One reusable prompt template that you'll use repeatedly.

### Why This Matters

The value of AI compounds when you systematize. One good template saves hours over months.

---

## Day 5: Beyond Text

**Goal:** Try a non-text AI capability.

**Time:** 30 minutes

Pick ONE:

### Option A: Image Analysis
Upload an image to Claude and ask questions about it:
- "What's in this image?"
- "Extract the text from this screenshot"
- "Describe what's happening here"

### Option B: Document Processing
Upload a PDF or long document:
- "Summarize the key points"
- "What are the main arguments?"
- "Find all mentions of [topic]"

### Option C: Code Understanding
Even if you're not a developer:
- Paste code and ask "What does this do?"
- Ask "How would I modify this to [change]?"
- Request explanations of technical documentation

### Today's Outcome

One completed task using AI beyond basic text generation.

---

## Day 6: Combining Tools

**Goal:** Complete a workflow using multiple AI capabilities.

**Time:** 60 minutes

### The Workflow

1. **Research** (Perplexity or Claude web): Find information on a topic
2. **Synthesize** (Claude): Combine and analyze the research
3. **Create** (Claude): Generate an output based on the synthesis
4. **Refine** (Claude): Polish through iteration

### Example: Creating a Presentation

```
Step 1 (Perplexity): "What are the top 5 trends in [industry] for 2026?"

Step 2 (Claude): "Based on this research [paste], identify which trends
are most relevant for [my company/audience]."

Step 3 (Claude): "Create a 10-slide presentation outline covering the
3 most relevant trends. Include key talking points for each slide."

Step 4 (Claude): "Expand slide 3 with more detail and add a compelling
opening hook for slide 1."
```

### Today's Outcome

One complete deliverable created through a multi-step AI workflow.

---

## Day 7: Reflection and System Setup

**Goal:** Establish your ongoing AI workflow.

**Time:** 45 minutes

### Part 1: Audit Your Week (15 min)

Answer these questions:
- Which tasks worked best with AI?
- Where did you still need heavy editing?
- What took longer than expected?
- What would you do differently?

### Part 2: Define Your Stack (15 min)

Based on your experience, decide:

| Task Type | Tool | Why |
|-----------|------|-----|
| Writing | [your choice] | [reason] |
| Research | [your choice] | [reason] |
| [Other] | [your choice] | [reason] |

### Part 3: Set Up for Week 2 (15 min)

- Save your best prompt templates
- Bookmark the tools you'll use
- Decide: upgrade to paid, or continue free?
- Identify one task to automate next week

### Today's Outcome

A documented system for how you'll use AI going forward.

---

## What's Next

After your first week:

**Week 2:** Go deeper on one tool. If you like Claude, explore Claude Code. If you like automation, try n8n.

**Week 3:** Build your first automation or template library.

**Week 4:** Evaluate what's working. Cut what's not.

See: [The AI Learning Trap (And How to Escape It)](/articles/fundamentals/ai-learning-trap) for what to avoid.

---

## Quick Reference: Daily Checklist

| Day | Task | Outcome |
|-----|------|---------|
| 1 | Complete one real task | Task done with AI |
| 2 | Iterate and refine | Polished content |
| 3 | Research mode | Brief with sources |
| 4 | Create template | Reusable prompt |
| 5 | Try non-text AI | Image/doc/code task |
| 6 | Multi-step workflow | Complete deliverable |
| 7 | Reflect and systematize | Documented approach |

---

## Common First-Week Mistakes

| Mistake | Fix |
|---------|-----|
| Spending day 1 "learning" instead of doing | Start with a real task |
| Accepting first output as final | Always iterate |
| Trying too many tools | Pick one, go deep |
| Over-engineering prompts | Simple, clear instructions work |
| Not saving what works | Document your templates |

---

## Next Steps

- [AI Glossary: 100 Terms You Actually Need](/articles/fundamentals/ai-glossary-2026)
- [Which AI Tool Should You Use?](/articles/fundamentals/which-ai-tool-decision-tree)
- [The AI Learning Trap (And How to Escape It)](/articles/fundamentals/ai-learning-trap)
- [Back to AI Fundamentals](/articles/fundamentals/ai-fundamentals-guide)

---

## Sources

- [Claude Documentation](https://docs.anthropic.com)
- [Perplexity](https://www.perplexity.ai)

---

*Last verified: 2026-01-20. Tested workflow with complete beginners.*