Well the argument is that B is no longer sustainable because of the scenario you describe in A. If you do a bunch of work on Navier Stokes but then OpenAI gets all the press, then what was the point?