📝 Egor Howell Blog

A Comprehensive Guide to Learning AI: Essential Resources and Advice from a 4-Year Practitioner

Artificial Intelligence (AI) and machine learning (ML) have rapidly evolved into some of the most exciting and impactful fields in technology today. After working in AI and ML for over four years, I want to share a curated list of resources—books, courses, and tools—that have genuinely helped me on my journey. Whether you’re a beginner or looking to deepen your expertise, this guide breaks down essential learning materials by categories: programming and software engineering, mathematics and statistics, machine learning, deep learning and large language models (LLMs), and AI engineering.

1. Programming and Software Engineering: The Foundation of AI Work

Strong programming and software engineering skills are indispensable for anyone aspiring to work in AI. Greg Brockman, OpenAI’s CTO, emphasizes this too.

• Why Python?
  Python remains the dominant language in AI and ML due to its extensive ecosystem of libraries and community support. Most AI infrastructure projects and machine learning jobs revolve around Python, and this trend is likely to continue for at least the next five years.

• Beyond Python
  The role of an AI engineer today often leans closer to software engineering than pure machine learning. Learning backend languages like Java, Go, or Rust (which I personally use) might be valuable for future-proofing your skills.

• Recommended Resources to Learn Python:

• Learn Python course by freeCodeCamp (4 hours, excellent for beginners)
• Python for Everybody specialization on Coursera (highly popular and well-reviewed)
• Practice coding and problem-solving on platforms like HackerRank and LeetCode
• NeetCode for learning data structures, algorithms, and system design fundamentals
• Harvard CS50 Introduction to Computer Science (best for total beginners wanting a solid computer science foundation)

Pro Tip: The best way to learn programming is by doing. Use these resources to grasp the fundamentals, then dive into hands-on projects immediately.

2. Mathematics and Statistics: Understanding the “Why” Behind AI

While some argue that deep math knowledge isn’t needed to use AI models (especially with pre-trained large language models), I believe that to become a top AI practitioner, you should understand the math under the hood.

• Key Areas of Math to Focus On:
• Statistics
• Linear Algebra

• Calculus

• Top Resources:

• Practical Statistics for Data Scientists: the best book for applied statistics in AI and ML, including hands-on Python examples.
• Mathematics for Machine Learning: focuses on linear algebra and calculus; it’s dense but invaluable.
• Mathematics for Machine Learning and Deep Learning specialization by DeepLearning.AI: tailored math courses focused specifically on AI and ML applications.

These three resources cover all essential math topics needed for a successful AI career without overwhelming you with irrelevant content.

3. Machine Learning: Building Your Core AI Knowledge

Machine learning is the backbone of modern AI, and gaining a strong understanding here is crucial.

• Recommended Books and Courses:
• Hands-On Machine Learning with Scikit-Learn, TensorFlow, and Keras by Aurélien Géron: the ultimate practical book covering fundamentals, coding, and advanced topics like reinforcement learning and autoencoders. If you get only one book, make it this one.
• Machine Learning Specialization by Andrew Ng (Coursera): a classic and highly effective course that teaches both theory and practical aspects using Python (recently updated from MATLAB/Octave).
• The 100-Page Machine Learning Book by Andriy Burkov: a concise overview ideal for reference and quick learning.

• The Elements of Statistical Learning: a traditional and theoretical text for deep understanding of statistical learning methods.

• Bootcamp Option:

• Zero to Mastery Complete AI, Machine Learning & Data Science Bootcamp: an intensive, project-based course perfect for those wanting structured learning and practical experience. The course is known for its strong community support and success in job placements at top companies.

4. Deep Learning and Large Language Models (LLMs): Understanding the Cutting Edge

Deep learning powers the latest advances in AI, including generative models like GPT and diffusion models.

• Why PyTorch?
  PyTorch is quickly becoming the de facto deep learning library, favored by researchers and used in over 77% of research papers (2021 data). Most Hugging Face models are PyTorch-exclusive, making it a must-learn tool.

• Recommended Learning Path:

• Start with PyTorch basics.
• Take the Deep Learning Specialization by Andrew Ng: covers CNNs, RNNs, and introduces LLMs.
• Watch Andrej Karpathy’s Introduction to LLMs video (1 hour): an insightful overview of generative AI and the current landscape.
• Follow up with Karpathy’s Neural Networks Zero to Hero course: an in-depth, hands-on course where you build a GPT model from scratch using raw numpy arrays—an excellent way to deeply understand neural networks.
• For a textbook, Hands-On Large Language Models by Jay Alammar (author of the famous “Illustrated Transformer” blog) offers an intuitive, up-to-date explanation of transformers and LLMs.

5. AI Engineering: Productionizing AI for Real-World Impact

Understanding AI concepts is great—but deploying models to production and building scalable AI-powered products is where real business value lies. AI engineering focuses on this crucial phase.

• What is an AI Engineer?
  Unlike ML engineers who often build models from scratch, AI engineers typically integrate existing foundational models (like LLaMA, Claude, ChatGPT) into products and infrastructure, handling deployment and scaling.

• Key Recommended Books:

• Practical MLOps: covers containerization, cloud deployment, and the practicalities of shipping ML models.
• AI Engineering by Chip Huyen: a highly praised book by a leading expert focused on AI/ML deployment best practices and hands-on examples.

Final Thoughts and Learning Philosophy

The resources above may seem overwhelming at first, but the key is not to consume everything at once. Pick a resource that suits your current level, dive deep, and build projects. Learning iteratively through concrete projects is the best way to truly master AI.

To help you keep motivated and on track, consider these three principles inspired by Andrej Karpathy:

1. Learn on demand, iteratively, by accomplishing concrete projects.
2. Teach or summarize what you learn in your own words.
3. Only compare yourself to the person you were yesterday—not to others.

If you want personalized guidance, coaching, or CV reviews tailored to your AI/ML career goals, I offer one-on-one coaching packages. Check the link in the description below to learn more.

Ready to Start Your AI Journey?

Whether you’re just beginning or looking to deepen your AI expertise, these resources and strategies will set you on the right path. Remember, consistent practice and real-world projects are your best teachers.

Happy learning and welcome to the exciting world of AI!

Useful Links (from the video description):

• FreeCodeCamp Learn Python course
• Coursera Python for Everybody specialization
• HackerRank and LeetCode coding platforms
• NeetCode algorithms and system design
• Harvard CS50 course
• Practical Statistics for Data Scientists (book)
• Mathematics for Machine Learning (book and course)
• Hands-On Machine Learning with Scikit-Learn, TensorFlow, and Keras (book)
• Andrew Ng’s Machine Learning and Deep Learning Specializations
• Zero to Mastery AI, Machine Learning & Data Science Bootcamp
• Andrej Karpathy’s LLM Introduction and Neural Networks Zero to Hero course
• Hands-On Large Language Models by Jay Alammar
• Practical MLOps (book)
• AI Engineering by Chip Huyen (book)

Embark on your AI journey today with confidence and the right tools at your disposal!