The AI Learning Adventure: Your Roadmap to Knowledge

AI, ML

Studying illustration

Two AI-related posts in a row? Well, hype is hype! One way to cut through it is to learn about the subject matter and use that knowledge to distinguish between what’s genuine and what’s merely a sales pitch. (I promise this is the last AI-related post this year)

Over the past year, I’ve invested a significant amount of time studying the field of AI, using a variety of resources as my guide. Today, I’m excited to share this list with you, so you can navigate the vast ocean of online content and find the best resources available. I’ll categorize the material into several areas, highlight the most important aspects within each category, and recommend resources in multiple formats - like articles, videos and books - tailored to different learning styles.

I hope you find this list helpful in figuring out what to learn. Happy studying!

Artificial Intelligence (AI)

What you should learn:

  • Definition of Artificial Intelligence (AI)
  • Types of problems AI can effectively address
  • Challenges and considerations associated with AI: security, ethics, and accuracy

Learning resources:

Machine Learning

What you should learn:

  • Definition and purpose of a machine learning model
  • Understanding features and labels in datasets
  • Introduction to feature engineering techniques
  • Distinctions between classification, regression, and clustering
  • Overview of neural networks and their functions
  • Introduction to model evaluation metrics (e.g., accuracy, F1 score, precision, and recall)

Learning resources:

Generative AI

What you should learn:

  • Definition and scope of generative AI
  • Capabilities for text, image, video, and speech generation

Learning resources:

Large Language Models (LLMs)

What you should learn:

  • Definition and purpose of large language models
  • Understanding transformer architecture
  • Overview of various large language models available
  • Guide to prompt engineering and its fundamental principles

Learning resources:

Local LLMs

What you should learn:

  • Instructions for running Ollama locally
  • Steps for downloading models
  • Techniques for invoking models via Command-Line Interface (CLI)
  • Guidance for invoking models through REST APIs
  • Setting up Open WebUI and integrating it with Ollama

Learning resources:

GitHub Copilot

What you should learn:

  • Setup procedures for integrating GitHub Copilot with Visual Studio Code
  • Utilizing chat mode for code generation
  • Exploring agentic mode capabilities for enhanced productivity
  • Best practices for effective prompt design
  • Understanding the limitations and ethical implications of AI-assisted coding

Learning resources:

Azure AI

What you should learn:

  • Creating and managing projects within Azure AI Foundry
  • Steps for deploying trained models
  • Testing models using the playground feature
  • Invoking models via REST APIs
  • Adjusting invocation parameters for specific use cases
  • Monitoring model performance with Azure tools

Learning resources:

Embeddings

What you should learn:

  • Definition and significance of embeddings in AI
  • Techniques for creating embeddings
  • Understanding vector databases used for embedding storage
  • Methods for querying embeddings effectively
  • Techniques for improving embedding quality

Learning resources:

LangChain

What you should learn:

  • Purpose and capabilities of LangChain
  • Instructions for invoking models using the LangChain SDK
  • Steps for creating and querying embeddings with LangChain SDK
  • Building a chatbot leveraging LangChain functionalities

Learning resources:

Retrieval - Augmented Generation (RAG)

What you should learn:

  • Overview of RAG and its components
  • Use cases for implementing RAG
  • Workflow of RAG in practice
  • Building a RAG system utilizing Azure AI Foundry

Learning resources:

AI SDK

What you should learn:

  • Creating UI for chatbot
  • Generating text response
  • Building chat prompt
  • Rendering streaming response

Learning resources:

Agentic AI

What you should learn:

  • Definition and role of agents in AI
  • Exploration of agentic capabilities
  • Practical use cases for different agent applications

Learning resources:

Model Context Protocol (MCP)

What you should learn:

  • Definition and meaning of MCP
  • Availability of MCP tooling
  • Steps for creating an MCP-driven API
  • Integrating MCP-driven tools with Ollama
  • Best practices for securing MCP APIs

Learning resources:

Agent2Agent (A2A) Protocol

What you should learn:

  • Definition and meaning of A2A
  • Core concepts: agent card, discovery, task lifecycle

Learning resources: