AI Terms Explained (for Web Developers)

1. AI (Artificial Intelligence)

Software that can make decisions or predictions like a human would.

  • Example: Gmail spam filter → decides if an email is spam.
  • Even simple rules + ML models count as AI.

2. Generative AI

A type of AI that creates new content instead of just analyzing.

  • Examples:
    • ChatGPT → generates text
    • MidJourney / DALL·E → generates images
  • Think of it as auto-complete on steroids.

3. LLM (Large Language Model)

The engine under the hood of most generative AI chatbots.

  • A neural network trained on massive text data.
  • Examples: GPT-4, Claude, Gemini.
  • For developers: like a super-powerful text API → input = prompt, output = response.

4. RAG (Retrieval-Augmented Generation)

Technique to combine search + LLM.

  • Flow:
    1. User asks a question
    2. System searches database/knowledge base
    3. Relevant info is passed to LLM
    4. LLM answers based on real data
  • Example: A chatbot that answers using your company’s FAQs/docs.

5. MCP (Model Context Protocol)

An open protocol (from OpenAI) for tools, apps, and AI models to talk in a standard way.

  • Think of it like REST for AI.
  • Instead of custom integrations, AI can connect with GitHub, databases, or Slack consistently.

6. Agentic AI (AI Agents)

Goes beyond simple Q&A → AI can plan and act.

  • Capabilities:
    • Break tasks into steps
    • Call APIs
    • Write & run code
    • Remember context
  • Example: “Book me a flight, pay for it, and add details to my calendar.”
  • For devs:
    • LLM = function call (input → output)
    • Agent = worker service that uses tools & APIs to finish a task.

✅ Quick Summary

  • AI → broad idea of machines acting smart
  • Generative AI → creates new content
  • LLM → the engine for text generation
  • RAG → search + LLM → answers from real data
  • MCP → protocol for AI ↔ tools integration
  • Agentic AI → AI that plans, uses APIs, and takes actions