Even if you don’t do ‘respect’; there’s managerial self interest to work with:
There are some industries that are in love with the notion of disruptive genius; super-achievers who are ten times as productive as the common rabble, etc. and in some cases there may even be an element of truth to it(one certainly hears terrible things about software development handled on the “have we tried using the cheapest programmers and sheer weight of numbers?” principle). In other cases, it’s less obvious that it’s true; but easy measurement makes it attractive(sales-heavy areas, say, offer potentially flawed but very handy metrics practically free for the taking); and in others it’s less obviously true but seems to have become part of the compensation structure(get tons of work for comparatively little money from the peons hoping to make partner by paying the partners well enough for it to seem worth the effort); and even when they aren’t ritually adhering to the "fire the bottom ten percent every year, just because’ rule; they tend to be willing to endure a fair amount of burnout, churn, and ugly politicking by selecting for heavily motivated, but hard to satisfy, people rather than the unambitious but reliable.
This isn’t always the irrational choice; but several things have to be the case for it to be rational: worth has to be reasonably measurable; or aggressive ranking is just flailing; there has to be a substantial performance gap between the great and the mediocre; or the very real costs of churn and recruiting will start to eat into the profits; and some sort of innate talent or extreme drive has to be worth more than experience; or the loss of institutional experience caused by the drive to recruit genius wouldn’t be worth it.
Again, it isn’t obvious that these conditions never hold; sometimes nothing but a rockstar will do; but I suspect that “personal assistant” is oh-so-very-much not a position of this flavor. Assuming that they aren’t so hideously incompetent or glaringly incompatible that it was a bad idea from the start; what’s a better bet: that the new one will use their genius to disrupt personal assisting and be super effective; or that the experienced one knows your tedious logistical details so well that you don’t even know how many of them get picked off and dealt with before you learn of them?
There are plenty of positions in a similar situation; where internal experience is valuable; the candidate pool isn’t better than the current employees(or not better enough to be worth the costs of disruption); and the value of being able to retain people by pretty much just offering them market rate and their preference for the status quo is greater than the value of getting some extra work out of people grasping for the brass ring; but expecting to either be lucratively rewarded or quit in bitterness after some years of trying. Also good for the positions of more trust than honor, where the reliable; but merely competent, are a great deal safer and more predictable than the highly motivated; but mercenary or too ambitious to be easily satisfied.
Aside from any being-a-total-jerk he might be engaging in; it sounds like Musk is getting confused about which of his employees fall into what category. Perhaps when hiring rocket surgeons; he is entirely correct that relentless weeding in the search for overachieving rockstars who value stock options is the best strategy; but that doesn’t mean that this is true of every position he wants to fill( see also: his ongoing problem with automobile manufacture on the theory that assembly line workers are really looking for an overworked startup culture with fairly mediocre compensation; but stock options!).
This isn’t to deny the existence of people who seem to have no useful skills except an uncanny knack for organizational survival; but the acolytes of “up or out!” consistently underestimate the value of the reliable and unambitious.