So compare: in one field, you start earning a median $65k at age 22 with only a bachelor's degree, and in the other, you start earning $65k in your mid to late thirties, after a bachelor's, a PhD, and often at least one post-doc. Everything before that is "low-pay grinding", except that the academic definition of "low-pay" ranges from lower-middle class ($25k-$30k for the better-paid grad-students and worse-paid post-docs) to taking on debt (the humanities).
A median means that half the CS grads made less than 65k. Tenure track also comes with some eventual benefits you won't receive with a bachelor's in CS. Let's be clear about what we're actually comparing.
This is false.
For example, X% could make exactly 65k and only the lower ((100 - X) / 2)% would make less.
...where X \in (0, 100]
However I also doubt that many CS grads at all are making significantly less than $65k. Those that are making appreciably less than that are almost certainly not working in the field, or they are not working in the US. I wager that the salary distribution of CS grads working in the field in the US is pretty damn tight.
Granted, the job market was crap and I make 6 figures now, but there are still plenty of sub 40k programming jobs out there and people happy to take them.
That's not how medians work
Your assuming salaries are not inflated and that the ratio of people reporting salaries is constant at all income ranges and the reported numbers are accurate. I doubt any of those are true. So, IMO to observe a 65k median suggests the actual median is below 65k.
PS: Don't forget grad students count as recent CS grads and there generally not making anywhere close to 65K.