The Unexpected Possibilities of Text-to-Text

Explore how AI transforms plain text, structured data, and code and unlock unexpected possibilities.

When to Use This Activity

Perfect for teams who are just getting started with AI and only have access to text-based AI tools. This activity helps people discover that even with just text capabilities, there's so much more they can do than they might think.

Especially useful when:

Teams feel limited by "just having text"
People aren't sure how to use AI beyond chat
You want to show the versatility of text-based AI
Teams need practical ways to work with data and automation

Key Learning Outcomes

Transform everyday text into powerful formats (CSV, JSON, code)
Use AI to work with data in ways you hadn't considered
Create simple automation using AI-generated code
Build confidence in exploring AI capabilities

Methodology Overview

Step	Title	Description	Cards Used
1	Opening Hook	Get participants to recognize their daily copy-paste tasks and show how AI can help	Text Input, Text Output
2	Plain Text Transformation	Demonstrate basic text reformatting using meeting notes example	Text Input, Text Output
3	Text to Structure	Show how to convert unstructured text (emails, notes) into organized data (CSV, tables)	Text Input, Structured Text Output
4	Structure to Text	Reveal how to turn dry data into engaging content	Structured Text Input, Text Output
5	Code Generation	Surprise participants by turning text descriptions into working automation	Text Input, Code Output
6	Hands-On Practice	Let participants experiment with their own real-world examples	All Previous Cards

Cards

These cards will help demonstrate the versatility of text:

Materials Needed

The mat
Access to an AI assistant for live demos
Optional: Excel for showing structured data examples
Optional: Pre-recorded demos as backup

Step-by-Step Guide

The power of this activity lies in the surprise - participants suddenly realize they can do things they thought needed special tools or coding skills.

Step 1: Opening Hook (5 minutes)

Start with a simple question:

Who here spends time copying and pasting stuff between documents?

After hands go up (they always do), follow with:

Let's see how AI can help with that - and do a lot more than just copying and pasting.

This gets everyone nodding - they all do this daily. Perfect setup for showing text transformations.

Step 2: Plain Text Transformation (15 minutes)

Set up the first card combination:

Set up the mat with Text Input and Text Output cards.

Start with something they're familiar with (basic text improvement) before revealing the more powerful transformations.

Begin with a common workplace scenario:

Let's start with something you might do every day - turning rough notes into different formats for different audiences.

Example: Take messy meeting notes like this:

"talked about q4 goals. mary will handle client outreach. tom needs to update sales figures by friday. need to schedule follow-up next week" Ask AI to transform it into:

A professional summary for your boss
A bullet-point action list for the team This shows how the same text can become two different useful formats - and it's just the beginning of what we can do with transformations.

Why This Works: This example resonates because:

Everyone takes quick notes
Everyone needs to "clean them up" later
Everyone needs different versions for different audiences

Step 3: The Structured Data Revelation (20 minutes)

Text to Structure (10 minutes)

Replace the Text Output card with Structured Text Output.

Ever spend time manually creating spreadsheets from emails or documents? Let's fix that.

Example: Turn Messy Text into Clean Data. Take this email:

"Hi team, here are the Q1 results: Marketing spent 45K, Sales 62K, and IT 38K. All departments were under budget except Sales who went over by 2K." Ask AI to create a budget CSV:

Department,Spent,Status
Marketing,45000,Under budget
Sales,62000,Over budget
IT,38000,Under budget

Try these transformations:

Meeting notes → Action items table
Product descriptions → Feature comparison matrix
Email thread → Status report spreadsheet

Structure to Text (10 minutes)

Now flip it - put Structured Text Input and Text Output cards on the mat.

Got spreadsheets full of data? Let's turn them into something people actually want to read.

Example: Turn Data into Insights. Take this sales CSV:

Product,Units,Revenue,YoY_Growth
Laptops,1200,960000,15
Monitors,800,240000,-5
Keyboards,2000,100000,25

Ask AI to write:

A quick summary for the team
A detailed report for stakeholders
Bullet points for your next presentation

The magic moment comes when people realize they can go both ways - from messy text to structured data AND from dry data to engaging content.

Step 4: The Code Surprise (20 minutes)

Switch to the Code Output card for the final revelation.

Think about those repetitive tasks you do. Now watch how we can turn text into actual automation...

Start simple:

First, write out what you want to automate in plain language
Then, ask AI to turn it into code
Finally, have AI explain the code so you understand it

Example: "I need to rename 100 files to add today's date at the start" Turns into a simple script you can actually use!

Focus on practical examples they can use tomorrow. Avoid complex coding scenarios.

Step 5: Hands-On Exploration (15 minutes)

Let participants experiment with their own examples. Some proven starting points:

Pick something you do repeatedly and let's turn it into:
A structured format you can import
A simple script to automate it
Or both!

Facilitation Tip: Walk around and help people refine their prompts. Often they just need a small tweak to get great results.

Step 6: Wrap-up (5 minutes)

End with this reflection:

Look at what we did - we turned plain text into structured data and working code. Tomorrow, when you catch yourself doing something repetitive, remember: you might be just one conversation with AI away from automating it.

Facilitation Tips

Keep it Practical: Focus on tasks people actually do
Handle Complexity: If someone suggests something too complex, help break it down
Build Confidence: Emphasize that they don't need to understand code to use it
Stay on Track: It's easy to get lost in possibilities - keep coming back to practical uses

What's Next

Close with this action-oriented message:

Tomorrow, catch yourself doing something repetitive. Stop and ask: "Could I ask AI to help automate this?" Start small - even saving a few minutes feels like magic.

Suggest these immediate applications:

Converting their next meeting notes into action items
Turning their task list into a structured format
Creating a simple script for a repetitive task

The goal is to have them try at least one text transformation within 24 hours.

Variations

Quick Demo Version (20 minutes)

Skip the hands-on parts
Focus on 2-3 impressive demonstrations
End with specific examples they can try later

Advanced Version

Use an Agentic Code Editor to build a simple app.

Resources

For trainers who want to learn more:

Blog Posts

Can AI Build Your Next App? - Deeper dive into AI's code generation capabilities

Trainer Guides

What is an LLM? - Foundational concepts to help answer participant questions

Example Library

Text to Structured Text Examples - Additional transformation examples to use in demos
AI Code Capabilities - More code automation examples for advanced participants

LLM - Inside the Black Box

Through roleplay, participants discover that an LLM is essentially a next-word prediction machine. A fun way to demystify AI.

LLM - Image, The Unexpected Possibilities

Explore how AI transforms images and unlock unexpected possibilities.