Back to Course3 / 10

Welcome to the Launchpad

What Am I Looking At? (Gemini Vision)

AI just grew eyes.

For years, AI was text-only. You could talk to it but not show it anything. That changed. Tools like Gemini, ChatGPT-4o, and Claude can now look at photos, screenshots, diagrams, charts, and handwritten notes — and tell you exactly what they see.

This unlocks a completely new category of problems you can solve.

IMAGE PLACEHOLDER
DALL-E 3 / Midjourney / Ideogram
A futuristic split-screen illustration: on the left, a person holds up their phone to photograph a complex diagram. On the right, a clean AI interface displays structured insights extracted from that image. Dark background, electric blue and teal UI accents. No text.

5 Things You Can Do With Vision AI Today

1. Identify Plants, Animals, Objects

Snap a photo of that mystery plant in your garden → ask "What plant is this and how do I care for it?"

Find a bug in your house → ask "What insect is this? Is it harmful?"

2. Read & Extract Text from Images

A menu in a foreign language. A handwritten note. A screenshot of a document you can't copy-paste from. Just upload it and ask "What does this say?" or "Translate this to English."

3. Analyse Charts & Graphs

Upload a chart from a report and ask:

  • "What trend does this show?"
  • "Summarise the key insight in one sentence"
  • "What's the biggest change between 2022 and 2024?"

4. Critique & Improve Designs

Upload a screenshot of your website, slide, or document and ask:

  • "What's wrong with this layout?"
  • "How would you improve the visual hierarchy?"
  • "Is this colour scheme readable for someone with colour blindness?"

5. Turn Physical Notes Into Digital Text

Photo of whiteboard brainstorm → "Turn this into a structured bullet list"

Handwritten to-do list → "Transcribe this and sort by priority"

Watch It In Action

VIDEO PLACEHOLDER
Runway / Sora / Kling
Screen recording (60–90 sec): Open Gemini (gemini.google.com). Upload a photo of a food label. Ask "What are the top 3 ingredients I should be aware of from a health perspective?" Show the response. Then upload a screenshot of a complex spreadsheet and ask "What's the main insight from this data?" Record in 1080p, dark mode, no background music.

Try It Now

  1. Go to gemini.google.com (free account)
  2. Click the image upload button (camera icon)
  3. Upload any photo — your lunch, a receipt, a page from a book
  4. Ask: "Describe what you see in detail" or "What useful information can you extract from this?"

You'll be surprised what it notices.

The Prompt Pattern for Vision

"[Here is an image of X]. Please [task]. Focus on [specific aspect]. Format your answer as [structure]."

Examples:

  • "Here is a screenshot of my email inbox. List the 5 emails that look most urgent based on subject lines only."
  • "Here is a photo of my whiteboard. Transcribe everything written on it and organise it into categories."
  • "Here is a chart from our quarterly report. What does this tell me about customer growth in the last 6 months?"

Key takeaway: Vision AI turns your phone camera into an analysis tool. Anything you can photograph, you can now ask questions about.

Up next: Lesson 4 — Party & Event Planning (AI as your personal assistant)