Signing in

You will be sent to MillerKnoll sign-in.

AI

Text, Slides, Images, Meetings — What Each Is For

Multimodal does not mean "throw everything at Copilot." It means picking the representation that matches the job: prose for nuance, slides for scanability, images for spatial layout, transcripts for who said what.

Lesson 1

Same Copilot, different rooms.

Word and Outlook excel at drafting and rewriting long text. PowerPoint excels at structure and headlines — not deep argument. Teams Copilot needs transcription for spoken content. Image-capable flows help when layout or visual reference matters.

Start with the modality where your friction lives. This lesson maps common MillerKnoll tasks to the right starting point.

Core principles

  1. Text-first tasks: emails, briefs, policies, summaries of written material — Word or Copilot Chat with pasted context.
  2. Scan-first tasks: leadership updates, client overviews, workshop agendas — PowerPoint or slide-oriented prompts.
  3. Meeting tasks: action items and decisions — Teams with transcription enabled; chat-only recap misses spoken content.
  4. Visual reference tasks: comparing a photo, floor plan, or slide layout — attach image or open deck in context.
  5. Wrong modality symptom: polished prose that ignores your slide structure, or slides that flatten nuance from your brief.

Prerequisite: Copilot Basics — daily apps

Check yourself

When recapping what was said in a Teams meeting, what does Copilot need?

Do this in Copilot

List three tasks you did this week. Label each with the best starting modality.

Paste this into Copilot Chat and work through it before moving on.

Pick the right starting modality

I need to [TASK] for [AUDIENCE]. I have: [list what you have — notes, deck, transcript, image, spreadsheet]. Recommend the best starting modality (Word, PowerPoint, Teams, or Chat with attachments) and why. One paragraph, no prompts yet.
Open Copilot →
  • Goal and context

Did you run this in Copilot? Mark complete when you have tried it.

Next lesson: When Text and Visuals Agree (and When They Do Not) →