The Wall Street Journal spent fourteen years testing stock pickers against darts, and fifty years of fund scoreboards say the professionals rarely beat the plain index after costs. The same belief is reassembling around AI: predictive language models read as insight engines, with no scoreboard at all. Same error, new tape. Here is the test that would settle it, and the reason its result will not.
A genre of social media post promises the prompts that replace McKinsey: the SWOT, the market analysis, the deck. Nine prompts, sometimes ten. Every list omits the same one, the prompt that checks whether the AI got any of it wrong. That omission is the tell. The deck was never the product. The checking was, and it still has to be human.
Bryan Walsh’s Vox piece says college is still worth it, and he is right that the panic is overblown. But his “yes” leans hardest on one number, the Fed’s 12.5 percent, which is the return for the students who finish. Count everyone who starts and the honest answer is not yes or no. It is that “is college worth it” is the wrong question.
Forbes ranked the “most versatile” college majors from LinkedIn data. But LinkedIn defines versatility as the share working outside their major’s top two industries, counted only among graduates employed full time who keep a profile. For most majors, it is praising underemployment and calling it a skill.
I audited the famous numbers that say college pays, the million dollars, the 12.5 percent. None of them survives as a real return on investment.
Palantir didn’t drop its standards when it dropped the degree. It started judging what people can actually do. Families weighing college should make the same move.