One could also use language like "a decent chance". But research has shown that people translate vague phrases like "a decent chance" into probabilities in inconsistent ways. For an ML researcher who is already used to dealing in next-token probabilities that aren't rigorously determined, just stating a probability estimate directly is very natural.

It's not meaningless. He said he believes it's 50% likely. It's a remarkably clear statement, and the probililty model is his brain.