Researchers at OpenAI claim to have achieved state-of-the-art results on various language tasks using a scalable system that combines transformers with unsupervised pre-training. This approach pairs supervised learning methods with unsupervised pre-training, an idea explored by others in the past. The team's results suggest this method is effective and may warrant further research on larger datasets. --- Why it matters: This matters to AI researchers because it demonstrates a potentially powerful technique for improving language understanding without requiring large amounts of labeled data. Source: https://openai.com/index/language-unsupervised