It's great to think about what you can do personally, but actually the NSF/congress did create this problem by choosing the financial structure for science in the US in the 40s-50s. That was done centrally and was not an organic extension of how science was previously funded. Many possible alternatives such as direct student funding like they do in Canada (advisors who are advisors not feudal lords), are imaginable. The best fit depends a lot on field. For example I'd say medical research is some of the most-marketable, and most medicine faculty choose academia for the right reasons.
Unfortunately in my benighted corner of science things are going the wrong way. The NSF actually just created a master's program for managing a sort of scientific software development for which there is no market. It did this because saying scientists care about education and commercialization "sounds good" on a grant proposal. This will do nothing but waste 50million, and 40 highly skilled human years of time. It's laughable to anyone in the field, but such are the incentives at NSF.
Unfortunately in my benighted corner of science things are going the wrong way. The NSF actually just created a master's program for managing a sort of scientific software development for which there is no market. It did this because saying scientists care about education and commercialization "sounds good" on a grant proposal. This will do nothing but waste 50million, and 40 highly skilled human years of time. It's laughable to anyone in the field, but such are the incentives at NSF.