Meta note: it's plausibly net positive that all the training details of these models has been obfuscated, but it's frustrating how much energy has been sunk into speculation on The Way Things Work Inside OpenAI.

I was never a fan of this perspective, and I still think everything should have been transparent from the beginning.