That training and qualification is only as good as the processes and standards being trained for and qualified on. We don't have those processes and standards t

That training and qualification is only as good as the processes and standards being trained for and qualified on. We don't have those processes and standards to train against (and frankly I'm not convinced we should or even can) for generic "software engineers". I have a number of friends who are PEs, and it isn't the training and certification process that differentiates their work from mine, it is that there are very clear standards for how you engineer a safe structure or machine. But I contend that there is not a way to write such standards for "software". It's just too broad a category of thing. Writing control software for physical systems is just very different from writing a UX-driven web application. It would be odd and wasteful to have the same standards for both things. I do think qualification would make sense for more narrow swathes of the "software engineering" practice. For instance, "Automotive Control Software Engineer", etc.