The concepts behind the technology reshaping every digital job — from how machines learn at all to the generative models now writing, drawing and reasoning.
This is the heart of “Getting Digital in the AI era”. Start with the foundations — machine learning, neural networks and deep learning — then follow the thread into today's frontier: generative AI, large language models and the Transformer architecture they are built on. Each concept below links to the courses that teach it, so you can move from understanding a term to actually learning it.
Machine learning is the branch of artificial intelligence in which a system gets better at a task by finding patterns in data, instead of following rules a programmer wrote out by hand.
A neural network is a model made of layers of simple connected units whose connection strengths are tuned, through training, until the whole network maps inputs to the outputs you want.
Deep learning is machine learning with many-layered neural networks that learn their own features directly from raw data, rather than relying on features a human picks out first.
Generative AI is a class of models that produce new content — text, images, audio or code — by learning the patterns of a huge body of examples and then sampling fresh combinations from them.
A large language model is a neural network trained on vast amounts of text to predict the next piece of text, which turns out to be enough to make it write, summarise, translate and reason in language.
The Transformer is a neural-network architecture that processes a whole sequence at once and uses an 'attention' mechanism to weigh how much each part should influence every other part.
Prompt engineering is the practice of writing the input to a generative model carefully — instructions, context and examples — so that it reliably produces the output you actually want.
Retrieval-augmented generation is a pattern that fetches relevant documents at question time and puts them into a model's prompt, so its answer is grounded in specific, current sources rather than memory alone.
Natural language processing is the field of AI concerned with getting computers to work with human language — understanding it, generating it, translating it and pulling meaning from it.
Computer vision is the field of AI that gets machines to interpret images and video — detecting objects, classifying scenes, reading text in pictures and tracking what moves.
Agentic AI is the step from models that answer to systems that act: software built around a language model that plans a task, calls tools and APIs, checks its own progress and keeps going across multiple steps with limited human intervention.
MLOps is the engineering discipline of getting machine-learning models out of notebooks and into production — versioning, deploying, monitoring and retraining them — so they keep working after the world stops looking like the training data.
Responsible AI is the practice of building and deploying AI systems whose outcomes you can defend — fair, explainable, private and accountable — treating 'should we, and under which safeguards?' as an engineering requirement rather than an afterthought.
Python is a general-purpose programming language prized for readable syntax and a vast ecosystem of libraries, which has made it the default language of data and AI.
Data engineering is the discipline of building the pipelines and storage that move data reliably from where it is created to where it can be analysed or fed to a model.
Closes the gap most aspiring AI engineers hit first — they can write a little code, but not the programming-and-maths groundwork that machine-learning material takes for granted. It walks from language basics to a working model in PyTorch, with data libraries in between.
About 52 hours
A guided tour of the field's main neural architectures in which you implement each one rather than just read about it — plain networks first, then convolutional models for vision, sequence models and Transformers for language, and finally generative models that synthesise new images.
About 50 hours
Walks the full investigative loop on real, untidy datasets — pose a question, gather and clean the data, explore it, then turn the result into a visual story an audience can act on. Each project lets you choose the dataset, so the work doubles as portfolio material.
About 43 hours
For developers who can already prompt a model and now need to ship one. It concentrates on the engineering that separates a demo from a product — selecting and adapting models, wiring them to your own data through retrieval, working across images and audio, and measuring whether the result actually holds up.
About 56 hours
A short, vendor-grounded course for people who need to be genuinely conversant with AI on AWS — not researchers, but developers, analysts and decision-makers who must pick the right service, stand a model up, and keep it responsible. The trade is breadth for usefulness.
4 weeks
An advanced, four-course ExpertTrack about the unglamorous half of AI: the plumbing. It covers feeding models with well-built pipelines and warehouses, grounding them in organisational knowledge through retrieval, running them reliably at scale, and orchestrating agents that act on their own.
About 8 weeks
