Ethan Mollick reports GPT-6 Pro handles one-shot complex calculation and reasoning well

Ethan Mollick says the GPT-6 Pro model performs impressively on one-shot problems involving complex calculations, references, or sophisticated ideas.

He says the model is not especially fast or pleasant to use, but gives an "ask one question of the great oracle" experience. This is a single post from Mollick on X, and the report includes no benchmark results or test details.