I agree with you, but that’s kindof beside the point. Open AI’s thesis is that they will work towards AGI, and eventually succeed. In the context of that premise, Open AI still doesn’t believe AGI would be winner-takes-all. I think that’s an interesting discussion whether you believe the premise or not.
If they're saying it's not 'winner takes all' it's cover to shield them from the fact that they aren't (or may not always be) in the top spot. They would not be saying that if they were undeniably the best - they'd say the opposite.
