AI MODELS & TOOLS

• ChatGPT
  General-purpose chat, coding, Q&A.
• Google Gemini
  Web-aware tasks, images; daily limits.
• Claude
  Strong reasoning/writing; good for longer answers.
• Grok
  Real-time model on X; capable, less filtered.
• Jan
  Run models on your own machine; review settings to keep data local.
• Hugging Face
  Try community LLMs and Spaces in the browser.
• ElevenLabs
  High-quality text-to-speech / voice.
• PDF Summarizer
  Upload a PDF → get a concise summary.

Agentic / Action-capable AI models

• Microsoft Copilot
  Web-grounded answers; can take multi-step actions.
• Gemini
  Calls tools in the background; moving toward agentic use.
• ChatGPT (current free model)
  Can browse/call tools in-chat, session-based.
• Claude
  Handles multi-step reasoning; agentic when given tools.

PROMPTING TIPS

1. Front-load current information. Tell the model to research or summarize the topic online first, before submitting your initial prompt.
2. Specify the desired outcome. Detail the goal, and specify the format.
3. Open the LLM in a single browser tab. Multiple tabs open to the same model can potentially confuse the model.
4. Initially provide full context. Names, sources, tone, length, audience.
5. Request the model self-review. Ask the model to iteratively evaluate and refine its own response.
6. Test where it runs. Test the result in the target environment and request changes/fixes based on the current context.
7. Manage your personal cognitive load. Periodic, short, non-cognitive breaks will reinvigorate mental fatigue.

Go deeper with these power-user tools

• Google AI Studio quickstart
  Try Gemini in the browser; explore models and prompting.
• OpenAI API quickstart
  Create a key, make a call, extend with tools.
• OpenAI Agent Builder
  Define agents that call tools and follow workflows.