> point is that the pie-in-the-sky promises are insufficiently novel

You have no data to judge novelty on. If you think AI rocket science, they’re flubs. If you don’t, they have a shot. Either way, I can’t see novelty being unsatisfied.