Synthetic Data in plain English.

Synthetic data is artificially generated data — text, images, code, or structured records — created specifically to train or fine-tune AI models, used when real-world data is too scarce, too expensive to label, or too sensitive (medical records, financial data) to use directly. A common approach: use a strong existing model to generate thousands of example question-answer pairs, then train a smaller or specialized model on that generated set. Synthetic data has become central to frontier model training since 2024, partly because the internet's easily scraped supply of fresh real text is running low relative to the data appetite of ever-larger models. The risk is "model collapse" — training a model too heavily on other AI's output rather than real-world data can amplify errors and homogenize outputs over generations. Used carefully, alongside real data rather than instead of it, synthetic data is now a standard part of how frontier labs train and fine-tune models.