Machine Learning, Deep Learning and LLMs: The AI Family Tree
Know what AI, machine learning, deep learning and LLMs mean, and how they fit inside each other.
These words are often mixed up, but they fit inside each other like nesting boxes. AI is the biggest box: any computer system that does something that looks intelligent.
machine learning is the part of AI that learns from examples, like the spam filter in the previous lesson. deep learning is the part of machine learning that uses big neural networks with many layers. That is what recognises faces, understands speech and drives most of today's AI.
An LLM (Large Language Model) such as ChatGPT, Claude or Gemini is one kind of deep learning: a huge neural network trained on enormous amounts of text. So every LLM is deep learning, all deep learning is machine learning, and all machine learning is AI.
Key takeaways
- AI is the big umbrella: anything that behaves intelligently.
- Machine learning = AI that learns from examples instead of hand-written rules.
- Deep learning = machine learning with neural networks that have many layers.
- LLMs = deep learning models trained on huge amounts of text to predict the next word.
Check your understanding
Go deeper (more technical)
Machine learning also includes methods that are not neural networks at all, such as linear regression (the house-price pattern), decision trees and their combinations ("random forests", "gradient boosting"). These are still widely used for table-shaped data, like prices, risks and sales forecasts.
Deep learning took over around 2012, when a deep network called AlexNet beat every other method at recognising images. Three things came together: big datasets, better training tricks, and GPUs, which you will meet in Track 3. LLMs arrived after the transformer design was published in 2017 (Track 2).