I've noticed that "18 months" is a very popular prediction target, across a wide variety of prognosticators. Close enough to seem urgent, and if what they predict comes to pass anytime in the next few years, they can claim "See, I was right, just a little off on the timing because <excuse>". If it doesn't happen, most everyone will have forgotten their prediction. And in the unlikely event they get called out, they have over a year of random events to find a plausible fall guy for their failure.