1) The system only trains PhDs for academic research.
2) A PhD does not guarantee an academic job.
PhD skills, particularly in technical fields, can be quite valuable in non-academic careers. To make that transition, though, students need to do the legwork themselves to gain relevant experience and contacts. (e.g., want to move into a programming gig? Do some open-source work and put it on github to build credibility.) Peter Fiske's "Put Your Science to Work" book is a good resource for navigating the culture shift.
1) The system only trains PhDs for academic research.
There's lots of research jobs in industry and government, and a PhD is excellent preparation for them. Who else but a PhD-trained researcher would know enough to be able to run the research programs at Intel or Genentech? Come to think of it, who but a PhD-trained researcher would know enough to be able to start a company like Intel or Genentech?
[1] http://www.technologyreview.com/blog/mimssbits/26607/?p1=Blo...
I'm not sure what it means to be "trained as a professor" rather than "trained as an industrial/government researcher", because I don't think I've ever been trained as either.
Certainly (and this is actually a problem) there isn't enough training of PhD students in key professorial skills such as "how to teach a class" or even "how to write a research grant" -- this is something which freshly-minted professors have to figure out on their own.
1) The system only prepares PhD graduates for continuing work in academia.
2) PhD does not guarantee a job.
The system is broken and I'm glad more people are realising it. This is major news coming from Nature.