Possible scenario. We can’t know about the internal motivations of OpenAI unless they tell us and I haven’t seen any statements from them outside the fact that they don’t care if it’s bad at math.