Training data
noun
The material used to teach an AI model its patterns. What is abundant becomes easy for the model to reproduce; what is missing, rare or repeatedly copied can be flattened or lost.
Also known as training corpus
Training data is the material from which an AI model learns its patterns: writing, images, code, recordings and other examples gathered before the model is asked to produce anything of its own.
The composition of that material leaves fingerprints. What appears often becomes easier to reproduce. What is rare, local, awkward or badly represented is easier to mishandle. The model does not know that the missing edge was important. It only knows what the data made available.
The book returns to training data because human work can flow back into the next machine, along with machine-made work nobody checked. That is where slop stops being merely annoying. At scale, today’s indifferent output can become tomorrow’s lesson.