Computers do not understand words, but they are good with numbers. Embeddings turn meaning into numbers.
Direct answer
An embedding is a list of numbers that represents the meaning of data such as text. Items with similar meaning get similar numbers, so you can measure how close two pieces of text are. Embeddings power semantic search, recommendations and retrieval in AI applications.
A simple picture
Imagine a map where each sentence is a point. Sentences about cooking cluster together, while sentences about football form another cluster. An embedding is the coordinates of each point, though real embeddings have hundreds or thousands of dimensions.
Where they are used
- Semantic search: find results by meaning, not exact words.
- Recommendations: suggest items similar to ones you liked.
- Clustering: group similar documents.
- Retrieval for AI assistants: find relevant passages to give a model.
How similarity is measured
Systems compare embeddings with a measure such as cosine similarity. A higher score means the meanings are closer.
Where Teyro fits
Teyro's The AI Stack course turns this into short lessons you can play in a few minutes a day. It has 100 lessons, and you can start free.
The bottom line
Embeddings are the bridge between human meaning and machine maths.
