Where to Start Coding

New to coding? A clear, beginner-friendly guide to choosing your first programming language and starting the right way.

Want to build websites? Learn these three in order, because each one builds on the one before it.

Backend Development

Next comes the code that runs on a server: storing data, handling logins and answering the requests your pages send. Start with one language; you don't need all six.

AI & Machine Learning

For AI, Python comes first, then preparing data, then the models themselves.

Or Explore Every Course

Not sure which path fits? Browse all courses and start anywhere:

Write your first web page

HTML is the text that describes a page: its headings, paragraphs, links, images and forms. A working page takes only a few lines, which makes it the quickest place to begin.

Make it look the way you want

CSS decides how that HTML looks: colours, fonts, spacing, and layouts that rearrange themselves for a phone or a wide screen. Every page you build from here on uses it.

Make it respond

JavaScript runs in the browser and reacts to what visitors do: clicks, typing, submitted forms, and data arriving from a server. With HTML and CSS behind you, you can see every change it makes to the page.

Pick a first server language

Python reads almost like plain English, so it is a gentle first language for the server. It is also the main language for data analysis and AI, so what you learn here carries over to both.

Store and query data

Nearly every back end keeps its data in a database, and SQL is how you ask it questions: find these rows, join those tables, change that record.

Server-side scripting

PHP runs on the server and sends back finished HTML. WordPress is written in it, and most web hosts run it without any setup.

Large applications and Android

Java is common in big company back ends and in Android apps. It makes you state the type of every value, which keeps large programs easier to reason about.

Games and .NET apps

C# is Microsoft's language for .NET web services and desktop apps, and the scripting language of the Unity game engine.

High-performance programming

C++ gives you direct control over memory and speed. It is the hardest language on this list, and it is used where performance matters most: game engines, browsers and devices.

Get comfortable with Python

Almost all AI work is written in Python. Be at ease with lists, dictionaries, loops and functions before you open an AI library.

What machine learning is

How a model learns from examples instead of rules, and the words the rest of the course relies on: features, labels, and training and test sets.

Prepare the data

Real data arrives messy. Clean it, fill the gaps and put it on a common scale with pandas and NumPy, before any model sees it.

Machine learning algorithms

Regression, classification and clustering with scikit-learn: how each algorithm works, and how to measure whether it works on data it hasn't seen.

Neural networks and deep learning

How a neural network turns inputs into a prediction and learns from its mistakes, worked through in plain Python first, then as TensorFlow and PyTorch write it.

Images and language

Put those networks to work on pictures and text: image classification, language models, and the transformer design behind today's chatbots.

Frequently asked questions

I have never written code. Where do I actually begin?

Pick one language, do the first three lessons today, and write something that prints your own name. The goal of week one is not understanding — it is proving to yourself that you can make the computer do a thing. Understanding follows use, not the other way round.

Do I need a computer science degree?

No, and plenty of working developers do not have one. A degree helps with the theory-heavy roles — compilers, graphics, machine learning research. For most jobs, what you can build and explain matters more than where you studied it.

How much maths do I need for programming?

For most web, app and scripting work: arithmetic and a little logic. That is genuinely it. Graphics, cryptography, machine learning and game physics need real maths, but they are specialisations you can choose later, not gates you pass through first.

How do I stay motivated when I get stuck?

Make the next step smaller. Being stuck usually means the gap between where you are and what you are attempting is too wide — so shrink it until it is boring, then widen it again. And read the error message properly; it is nearly always more helpful than it first looks.

How long should I study each day?

Thirty focused minutes daily beats four hours once a week, because programming is a skill you build by repetition and it decays quickly between sessions. Protect the streak over the session length.