Read This to Know What Books to Read!

If you're new to the world of AI you're probably confused about where to start. Should I learn statistics? Should I code? Should I be taking calc 101? The answer to all these questions is yes... kind of. There's a specific order I'd suggest tackling these subjects in. The reason for this order is that it introduces widely applicable topics first before getting into the specifics. Even if you're already knowledgeable in AI, this pathway will be refreshing and interesting:

• First learn how to use your computer's command line - like really use it. Read Unix for Poets to get comfy exploring your computer and doing basic work. You'll find that many of the tasks you'll be tempted to write a program to solve can be addressed with just a few terminal commands. If you're on a Mac or Linux machine, this will make sense. If you're using Windows, I suggest exploring the UNIX-like world.

• Now you should learn how to code. Go through the Learn Python the Hard Way tutorial and build some stuff. Whenever you learn a new skill from this tutorial, find something to apply it to. Start working on little projects and then build up to more advanced things. Set goals that slightly push your boundaries but won't overwhelm you. This "goal setting" process is the best way to learn, but might require some mentorship, so reach out to someone who knows how to code and can help you set up your goals.

• It's really important to run your code online. It's not very useful to just have software sitting on your own machine. Learn about Flask so that you can build websites and other applications. Most projects you work on - whether it be a chatbot, NLP system, or ML model - should be put to the web. Also figure out how to deploy your apps on the cloud. I wrote a tutorial about that. Use Github!

• Learn about Machine Learning and AI. Start simple with some NLP projects. This book introduces you to a lot of the tasks that people in the field of NLP have to deal with. By this time, you'll recognize some repeated themes from Unit for Poets and some coding principles repeated from Learn Python the Hard Way. Then you can graduate to Hands-on Machine Learning, which, as the title suggests, is a solid hands-on guide to ML, deep learning, and inferential statistics in general.

• Dive deeply into the AI world by reading Speech and Language Processing, a book that covers all the major topics in the field. By this time you'll probably already understand how to program and the basics of AI, and this book will really hammer in the details and give you context. If you're interested in the mathematical underpinnings of these topics, read this book.

• Figure out how to set up a standard data science development environment by reading this blog post. This will show you how to do normal data science work - which was introduced in Hands-on Machine Learning.

• For job interviews, you might be asked to solve standard software-engineer interview problems - the dreaded "algo interview". Learn the basics of CS algorithms with this book then read the industry standard Cracking the Coding Interview and practice on Leetcode. ML/AI algorithms are not the same as CS algorithms, but some of the basic concepts of CS algorithms, like complexity analysis, are useful when analyzing ML algorithms too; and some ML algorithms, like HMM's use CS algorithm components (i.e. dynamic programming).

If you follow this path, you'll have the raw knowledge to be successful in the AI industry - as a data scientist, data engineer, machine learning engineer, or even an analyst. Good luck, and happy reading 😊