OpenAI's latest model, ChatGPT-o1-preview, is being tested for its ability to solve complex academic problems. The model has demonstrated proficiency in tackling financial math problems faced by venture capitalists and solving PhD-level physics problems from standard textbooks such as Jackson's Classical Electrodynamics. There is ongoing debate about whether the solutions might be part of its training data, but the model's capability to compress and solve these problems is notable. Additionally, it has been tested with private exam questions.