Exploring Computer Vision
🎯 Objective:
Discover how combining images with questions can unlock powerful AI assistance for everyday problems and presentations.
🛠️ Tools:
- Multimodal AI Assistants:
💡 How Others Used This:
- "I uploaded a photo of my coffee carafe and got exact measurements for the perfect coffee-to-water ratio."
- "Showed AI my spreadsheet of oddly formatted names and got a formula to extract just the first names."
- "Snapped a pic of a confusing parking sign and got a clear yes/no about whether I could park there."
- "Uploaded my presentation slides and got suggestions to improve design, flow, and key talking points."
📝 Challenge:
Today is about exploring real-world applications of image + AI assistance. Try any of these ideas or explore multimodal capabilities on your own:
Everyday Mysteries:
- Snap a photo of a mysterious household item or tool and ask what it's for.
- Troubleshoot a confusing error message on your device.
- Take a picture of a warning light in your car and get immediate help.
Document & Data Help:
- Upload handwritten notes for transcription or translation.
- Share a screenshot of a spreadsheet for formula help.
- Photograph a form you need help filling out.
Visual Problem Solving:
- Photograph a parking sign and ask about restrictions.
- Snap a repair/assembly problem you're stuck on.
- Share a screenshot of a software issue.
Presentation Assistance:
- Upload slides or images for a presentation and ask for feedback on:
- Design improvements
- Flow and structure
- Key talking points
📢 Example Prompts:
For Unknown Items:
I found this in my kitchen drawer [image].
1. What is this tool called?
2. What is it typically used for?
3. Are there any interesting facts about it?
For Error Messages:
I keep getting this error [screenshot] when I try to save my document.
1. What does this error mean in simple terms?
2. What are the most common causes?
3. Can you suggest step-by-step troubleshooting steps?
For Presentation Slides:
Here are my presentation slides on [topic] [image].
1. Can you suggest improvements to the design and layout?
2. Are there ways to make the flow of information clearer?
3. What key talking points should I emphasize?
💡 Tips:
- Ensure images are clear and well-lit.
- You still need to include relevant context in your questions.
- Try follow-up questions if the first response isn't quite what you need.
- Experiment with different ways of asking about the same image.
🤔 Things to Consider:
- Some AI assistants might decline to continue the conversation when you upload images with people.
- Personal or sensitive information should be masked in screenshots.
- Complex technical issues might still need professional assistance.
- Image quality significantly affects the AI's ability to help.