TL;DR - You don't need every tool. Learn the categories, keep one general assistant as your default, and add a specialist tool only when a task demands it.
Why it matters
The tool list is overwhelming and changes weekly. Chasing tools wastes time. Knowing the category for a task removes the paralysis - you'll always know where to start.
The map
- General assistants - ChatGPT, Claude, Gemini. Your default for writing, thinking, Q&A.
- Research - Perplexity, NotebookLM. Synthesize sources with citations.
- Images - Midjourney, DALL-E, Ideogram. Generate visuals.
- Meetings/notes - Otter, Fireflies, Granola. Transcribe and summarize calls.
- Coding - Cursor, GitHub Copilot, Replit. Build and edit software.
- Automation - Zapier, Make, n8n. Connect apps into workflows (Level 4).
Worked example - match the task
"Summarize this contract" -> general assistant (Claude is great with long docs).
"What do experts say about X, with sources?" -> research tool (Perplexity).
"Turn this call into action items" -> meeting tool, or paste the transcript into an assistant.
"Make this run automatically every day" -> automation (Zapier/n8n).
Steal this - your starter stack
Default assistant: one of ChatGPT / Claude / Gemini (pick one, learn it deeply).
Add when needed: Perplexity (sourced research), a meeting tool, Zapier (automation).
Rule: master one general assistant before collecting specialists.
Common mistakes
- Tool-hopping instead of getting good at one assistant.
- Using a chatbot for sourced research when a research tool would cite real links.
- Paying for ten tools you use once.
Good to know
The three general assistants are close in quality and trade leads often. Claude is loved for writing and long documents, ChatGPT for its huge feature set and data analysis, Gemini for tight Google Workspace integration. Any of them is a fine default - the skills you build here work across all three.