AI Vocabulary · 01 The Model
What is Synthetic Data?
Training examples generated by a model or a program instead of written or recorded by people.
When real data is scarce, expensive or private, teams let a model write the examples instead: questions and answers, conversations, code, rare cases. Those examples are then used for training.
It is already standard in post-training. NVIDIA reported in 2024 that over 98 percent of the data used to align its Nemotron-4 340B model was synthetically generated. NVIDIA sells chips and models, so it has an interest in the approach working.
The risk is a loop. Ilia Shumailov and colleagues showed in 2023 that training models again and again on their own output makes them forget the rare cases at the edges of the original data. They called it model collapse. Their finding is about repeated training on generated output, not about a carefully curated mix of real and synthetic data.
Why it matters for you
Synthetic data can fill gaps, cover rare situations and protect privacy. It also inherits the mistakes and bias of the model that produced it. Always check results against real examples with evals.
Own it
If you generate examples for your own model, through fine-tuning or distillation, keep real data as your reference, label what is synthetic, and check whether the provider's terms allow using its outputs for training.
Where it comes from
There is no single inventor of the idea. The term model collapse comes from The Curse of Recursion by Ilia Shumailov and colleagues (2023). The 98 percent figure is from NVIDIA's Nemotron-4 340B Technical Report (2024).
Related
Your AI Learning Hub
Learn the whole AI Vocabulary, step by step, and build with it. Build it. Own it.