What is possible with AI. Ten input and output combinations. Hands on the modes.
Explore what GenAI can actually do.
This module is about exploring what Generative AI can actually do across a range of everyday tools and formats.
Rather than reading about features. You will test them. You will use text, voice, images, code, files and web searches. Sometimes as input. Sometimes as output. Each task is quick. Together they reveal the variety of ways you can interact with GenAI.
The goal is to understand each mode. And to start thinking about how different combinations can support real work, shape smarter workflows and lay the foundation for building tools like agents.
Why we are doing this.
You are not just learning what GenAI can do. You are beginning to ask. Where could this help in my work. How could I use this differently to get better results.
Build hands-on familiarity with how GenAI works across different input and output formats.
Encourage strategic thinking about which combinations are most useful for different types of work.
Stimulate ideas for how GenAI can improve decision making, reduce effort or enhance outcomes.
Lay the groundwork for building customised workflows or agents aligned to your role or team needs.
By the end of this module.
Three things you will be able to do.
Identify and test a range of GenAI input output combinations using everyday tools. Text. Voice. Image. File. Code. Web.
Recognise how different modes can be combined to support real workflows, projects or decisions.
Generate early ideas for practical use cases or workflows where GenAI could improve speed, clarity or quality of outcomes.
Ten combinations. One thread.
We are going to work through every one of these in a single AI conversation. Each is a different way to put information in and get something useful out. By the end of this session you will have tested all ten.
Two heads-ups.
A small convention you will see in every prompt. And a real-world reality you will hit. Worth naming both before we touch a keyboard.
You will see // at the start of every line in every prompt today. It is not magic. It is not a programming command. It is how I organise my prompts so each idea is on its own line. AI reads it the same as a normal sentence. Use it. Or do not. Your choice.
Your IT team has shaped your Copilot. Some of the ten combinations may be turned off on your tenant. Voice-to-voice. File uploads above a certain number. Image generation. Web search. If you cannot find a feature. It is not you.
Name your chat.
A small habit. Big payoff. AI keeps every conversation you have. The chats with names are the ones you can find again. The ones without names are the ones you lose. Start naming yours today.
Open one AI thread. Paste this in.
We are going to work through all ten combinations in a single conversation. Your AI tool remembers what you have done across the thread. That is part of what we are testing. Open Copilot or your tool of choice now. Paste this prompt to set the scene. Then move to the first combination.
Text → Text.
Function
- Generates, extends, summarises or problem-solves using only text
- The most common and versatile input-output mode
- Ideal for drafting, exploring ideas and asking for step-by-step support
Possible Applications
- Drafting documents or communications
- Brainstorming strategies or workarounds
- Outlining plans or step-by-step instructions
Reflections
- What is one task in your role you could brief AI on this way?
- Where would this save you the most time in a typical week?
Text → Image.
Function
- Turns written ideas into visual outputs
- Moves you from abstract thinking to concrete visualisation
- Helps quickly communicate a concept or story
Possible Applications
- Concept art, slides or diagrams
- Mockups for proposals or campaigns
- Visioning and storytelling tools
Reflections
- What in your work might land better as a visual than as more words?
- Where do you currently wait on design support or settle for stock images?
Text → Code.
Function
- Translates natural language into working code or scripts
- Enables non-coders to prototype tools and workflows
- Useful for technical enablement and automation
Possible Applications
- Creating dashboards, calculators or automations
- Enhancing data workflows
- Rapid prototyping or learning technical concepts
Reflections
- Where in your role does a small bit of code or a simple page sit between you and what you need?
- What might you build for yourself or your team that you cannot today?
Text → File.
Function
- Converts AI-generated content into downloadable formats
- Supports creating shareable assets directly from chat
- Useful for packaging outputs like reports, spreadsheets or scripts
Possible Applications
- Exporting code, summaries or data into a ready-to-use file
- Creating templates or documents from scratch
- Sharing work outputs with colleagues or clients
Reflections
- What documents do you produce regularly that could come out of AI as a finished file?
- Which file format would save you the most rework if AI generated it directly?
Voice → Text.
Function
- Captures thinking, brainstorming or dictation via speech
- Supports hands-free interaction
- Great for working through ideas in real time
Possible Applications
- Voice notes turned into action steps
- Quick idea capture
- Low-friction input for those who prefer speaking over typing
Reflections
- When in your day would speaking to AI work better than typing to it?
- What captures or notes do you currently lose because typing is too slow?
Image → Text.
Function
- Interprets and explains uploaded visuals
- Acts as a visual analyst, coach or simplifier
- Translates images into useful written insight
Possible Applications
- Clarifying complex diagrams or charts
- Reviewing presentations
- Summarising visual data
Reflections
- Where in your role do you have to interpret a chart, slide or screenshot for someone else?
- What complex visual could you upload tomorrow and ask AI to translate?
Web → Text.
Function
- Searches the web for current, real-time information
- Acts like a research assistant or briefing engine
- Useful for updates, summaries and environmental scans
Possible Applications
- Market scanning or competitor research
- Prep for meetings or presentations
- Staying informed on emerging topics
Reflections
- What ongoing scan or briefing in your role could AI prepare for you?
- What sources would you tell AI to prioritise or avoid?
File → Text.
Function
- Reads, summarises or analyses documents and structured data
- Works well with PDFs, spreadsheets or long-form content
- Quickly surfaces insights or summaries
Possible Applications
- Summarising reports or policies
- Extracting insights from HR or project data
- Spot-checking for errors or trends
Reflections
- What document on your desk right now could you upload and ask AI about?
- What routine analysis takes you hours that AI could surface in minutes?
Voice → Voice.
Function
- Enables full conversational interaction with AI using speech input and spoken output
- Supports hands free, natural dialogue
- Useful for exploring ideas in real time or simulating conversational coaching, briefing or thinking aloud
Possible Applications
- Talking through decisions or priorities
- Exploring options or generating ideas in a coaching-style dialogue
- Accessibility support or multitasking input and output
Reflections
- What in your role would benefit from talking to AI like a thinking partner?
- Where do you do your best thinking out loud? Could AI sit in that conversation?
Code → Text.
Function
- Explains code in plain language
- Helps with reviewing, learning or collaborating across technical boundaries
- Makes code more accessible for non-developers
Possible Applications
- Code review
- Learning technical skills
- Debugging or improving existing tools
Reflections
- Where in your work do you see things written in technical language you need translated?
- Who around you writes code, formulas or scripts you sometimes need to interpret?
Close the thread. Take the summary.
Go back to the same AI thread you have been working in. Paste this prompt. Your AI has the context of everything you have just done. It will ask you what stood out. Answer honestly. Then it will offer a clean summary of the ten combinations as a take-away resource. Save the chat or copy the summary.
What are you taking back.
Which combination stood out for you. Where in your role could it fit. What is the first thing you would brief AI to do.
About two minutes each. Which combination from today stood out and why.
Name a specific place in your work where this combination could land. Be concrete.
Sketch the prompt. What would you tell AI to do. Bring it back to the room.
What you have done.
More than a tour. You now have working knowledge across every mode AI offers. And the start of a habit you can carry into the rest of the day.
Text. Image. Code. File. Voice. Web. You have used every mode AI offers.
For every combination you named where it could land in your specific work.
A single conversation that holds the whole session. Your AI knows the context now.
A clean structured resource of the ten combinations. Yours to save and refer back to.
A small habit with a big payoff. The conversations you name are the ones you can find again.
Your peers brought back the combinations they would try first. You leave with their picks too.