People often try to get into top tier high schools to help them in their STEM careers starting at ages 8-12. At elite universities, most students have already seen calculus by the time they get there. A student doing average at age 12 probably won't see calculus by then.
I think in reality, preparation for STEM takes years. It helps to have an education that fosters interest at a young age as well. I am specifically questioning the priorities of the US's view of education. As I said, we rely heavily on H1-B and have very large foreign populations in STEM degrees. Social media may not be the cause. What matters is that there clearly isn't early interest. Early interest translates to having more graduates. As I posted elsewhere, there is a clear difference in what degrees foreign students pursue in the US (more STEM) versus what US students pursue. This seems like a difference in values that I think we should examine.
Sure, children going to the most competitive universities will probably have began prep work before the age of 8. I would guess that "Did my parents go to an elite university?" has far more predicting power for who will enter an elite university than "What was my dream job at age 10." Elite schools are designed to be attended by only a small fraction of the population, hence the term elite. For the rest of the population, going to a state university, community college, or even a technology trade school can lead to meaningful employment.
I don't think you can move the extremes without moving the mean. We are already seeing the effects. First and second generation immigrants from foreign countries (like China) outperform Americans. This cycle could become self fulfilling. People expect immigrants to achieve at the extremes and prioritize other things in their lives. As you said, let's suppose having parents who go to these schools predict who will enter. At this point, recent immigrant families would outperform and others under perform and this can continue as cultural values are passed on. Is this is a sustainable situation? No, I think it bifurcates it and creates permanent classes.
On the other hand, more focus on STEM education broadly would translate to higher enrollments and self reliance. This also assumes that out performing immigrant families assimilate. Non-assimilation would exacerbate the bifurcation.
I think in reality, preparation for STEM takes years. It helps to have an education that fosters interest at a young age as well. I am specifically questioning the priorities of the US's view of education. As I said, we rely heavily on H1-B and have very large foreign populations in STEM degrees. Social media may not be the cause. What matters is that there clearly isn't early interest. Early interest translates to having more graduates. As I posted elsewhere, there is a clear difference in what degrees foreign students pursue in the US (more STEM) versus what US students pursue. This seems like a difference in values that I think we should examine.