It feels like this just horse-shoes one step closer to writing in actual programming languages again (Djikstra's "natural language is foolish" argument). Or is

It feels like this just horse-shoes one step closer to writing in actual programming languages again (Djikstra's "natural language is foolish" argument). Or is the argument here that this is an irreplaceable "sweet spot" for spec-writing?