What would be so hard to explain?

That OpenAI heard about the result and decided to throw a lot of money at it knowing it was within reach ?

That the model took an approach that was published years ago ?

That's a fake explanation, it skips explaining/justifying how OpenAi "heard about" the result

	▲	famouswaffles 11 minutes ago \| parent [-]
		We are told in the statement that there were rumours already going around about a Stokes result (I even know about the rumors in question. It was all over twitter in the right spaces) and that Tristan contacts OpenAI about the rumors. In the message, it's pretty clear Tristan has made some result. So either the rumors or Tristan's contact would explain it fine.