Item — HN Reader

> We could probably repurpose AP exams to fill a similar admissions role to A Levels

They are already being used like that in college admissions today.

> possibly use the AMC/AIME/USAMO more explicitly for admission

Already in use explicitly at most of the top CS programs today

> I do not think declining admissions rates at elite universities reflect that the students are any smarter or more prepared than they were 20 years ago, but rather they are much more cutthroat about many things that are orthogonal to being successful adults

The cutthroatness is what we should be optimizing for long term. Competition is what begets innovation.

Also, even at solid middle tier universities like UC Riverside, the calibre of student has increased dramatically over the past decade [0]

[0] - https://news.ycombinator.com/item?id=39369724

1 comments

> The cutthroatness is what we should be optimizing for long term. Competition is what begets innovation.

The difficult pill that society refuses to swallow is that most of the simple innovations that are feasible under our constraints have already been innovated. There are no more individual 'inventors'. The source of modern technological innovation are lucky discoveries that manage to scrape from the bottom of the easy-to-find barrel, genuine one in a million genuises and (mostly) the hard work of large groups over many years. Even the brightest top 0.1% of people mostly won't be coming up with ideas that a thousand people before them didn't already have.

Yet people will not settle for this. Instead of acknowledging it, we'd rather apply the good old human model of 'if 1x of something is good, then we need 1000000x of that' and take healthy competition to transform it into a ruthless race to the bottom.

Past a certain point of competition, you stop selecting for the metric you care about (academic performance, curiosity, creativity) and start selecting for people who are willing to slave away sunrise-to-sunset to work on tasks they know are almost meaningless but that contribute to your selection criteria. The metric becomes the only goal. You're also indirectly selecting for wealth, because rich parents can afford to spend lots of time with their child, not make the child work, pay for all their extracurriculars and tutoring.

The diminishing returns are immense, and we can already see that in countries that never fought this race to the bottom. Their students are more miserable than ever, and for what? Frankly, forcing children to waste away their best days to make them clear arbitrarily rising barriers that we know are meaningless is unthinkably heartless.