This is very confusingly presented.

https://gizmodo.com/leaked-documents-show-openai-has-a-very-clear-definition-of-agi-2000543339

If we were discussing the mystery of death, or even just the medical ontology of it, and I told you there was a big breakthrough that finally settled what we mean by death — there is death when a body is completely consumed by fire — you’d scoff that I misunderstood the exercise.

To say that X is necessary for Y, so we can use Y to be in mutual agreement that X has happened, is fairly common. It doesn’t tell us the definition of X or the nature of X or an analysis of X or any model or conception of it (beyond that entailed by the Y claim).

And that’s true when Y is both sufficient and necessary for X to obtain. In the case at hand, as conveyed in the article, it’s merely sufficient. So it’s even weaker.