Image generation produces visual content from descriptions or other inputs. In IT, it supports practical work such as diagrams, charts, mockups, logos, and technical documentation, not only artistic images. section 1.4.2
Use image generation for:
- Creating simple logos and presentation artwork. section 2.2.1
- Writing Mermaid or PlantUML markup for flowcharts, sequence diagrams, and architecture diagrams. section 1.4.2
- Producing HTML/CSS mockups when exact text and layout control matter. section 1.4.2
Logo prompt
Prompt: “I want a simple cloud that looks relaxed and relieved like the relieved emoji. Also add a similarly simple yellow sun in the background.” DALL-E produced a logo that closely followed this description. Review generated results, because creative tools may not reproduce an image exactly or follow every instruction. section 2.2.1
Table 1.3 AI diagramming capabilities and limitations
|
Need
|
What AI can do
|
Current limits
|
Better options
|
|---|---|---|---|
|
Network diagrams
|
Create Mermaid/PlantUML markup with tools like ChatGPT and Claude
|
Need extra steps to get Visio files
|
Try AI plugins for Visio or scripts that make VSDX files.
|
|
Infrastructure maps
|
Write templates for Terraform or Mermaid
|
May need manual fixes for complex maps
|
Use Lucidchart AI to help with map layouts.
|
|
System architecture
|
Help with PlantUML models
|
Basic graphics only
|
Try Figma with AI plugins.
|
|
Database schemas
|
Create SQL and ER diagrams
|
Limited visual options
|
Use a tool like Workik or ChatGPT for schemas.
|
|
UI mockups
|
Make HTML/CSS or markdown designs
|
Not polished enough for demos
|
Try Uizard or Figma’s AI tools for quick UI designs.
|
Compare AI capabilities and limits for network diagrams, infrastructure maps, system architecture, database schemas, and UI mockups. section 1.4.2
