September 12, 2024: OpenAI Unveils Self-Fact-Checking AI Model: o1 - OpenAI launches o1, a generative AI model that fact-checks itself. Available in two versions, o1-preview and o1-mini, it uses reinforcement learning for enhanced reasoning. Despite its high cost and some limitations, o1 excels in complex tasks like legal analysis and programming. Its currently subscription-based, with wider access planned. Critics note occasional errors and hallucinations, but its performance in various benchmarks shows promise.
