On the inevitability of career change

I've caught a bad case of Zeitgeist from twitter as a result of a number of recent posts on the difficulties PhDs face in getting jobs in academia, and on the continuing failure of any institutions (universities, departments, funding agencies) to do anything about it. My position on this is, as they say, nuanced. On the one hand, the discrepancy between number of graduates entering PhD programs and number of jobs available is indeed crazy. The refusal of PIs to acknowledge this in any way is nothing less than personal cowardice. The fact that, as far as the NIH (which pays my salary) is concerned, I don't exist is, quite frankly, unconscionable. And yet, on the other hand, I can't help but feel that, well, career changes are difficult for everyone. When I look at my friends outside the academy (yes, I still have some, more on that later), many of them were whacked in the face by the 2008 financial crisis just as their careers were getting underway, and many of them had to deal, quite suddenly, with career changes. If you think for a second that the banks, management consultancies or accountancy firms that had hired them invested heavily in helping them retrain for new careers just as they were keeping them in fear of redundancy for about three years, think again.
Going back further, neither my brother (degree in civil engineering in the late 90s, now works in strategic marketing (whatever that is)), not my sister (law in the early 2000s, went into PR and is now a civil servant at the ministry of justice), have had careers that were related to their studies or professional training. Heck, even my mother had an alt-ac career back in the late sixties (she dropped out of her PhD program for lack of funding and retrained as a secondary school history teacher). For none of them was any formal assistance in career change provided. For all of them it was bloody hard. To put it another way: if you go into the graduate training program at Goldman Sachs and decide after five/six years that investment banking is not for you (and I know a fair number of people with career paths like this), you wouldn't expect Goldman Sachs to help you change career. And let's be honest, at this point, most major research universities are about as warm and cuddly as the aforementioned octopus of global finance.
Of course, the error with this analogy is that Goldman Sachs doesn't routinely take on an order of magnitude more graduate trainees than it can find places for (although that is exactly what ended up happening in 2008). And to the degree that the majority of freshly minted PhDs can now expect not to be professors, there is clearly a bigger onus on graduate school (in so far as it constitutes professional training) to address this reality. Then again, how do you do that when no-one in your program has any experience of anything else?
Which gets me (in a rather long winded way) to my main point. As dangerous in my mind as continuing to support the status quo in terms of number of PhDs granted is the continuing infantilisation to which graduate students are subjected during the course of their PhDs. The view maintained that grad school is a continuation of studies rather than an equivalent to the aforementioned Goldman Sachs graduate training program doesn't lead to the kind of professional development mindset that will prepare students for the variety of career choices that come their way, and doesn't encourage PIs, universities, journals or professional societies to put forward genuine opportunities for professional growth to graduate students.
In essence, part of the preparation for the difficult career decisions  grad students will have to face is emphasizing to all involved that grad school is a career decision, not the delaying of one.