We asked five questions about the digital SAT — how it works, what the adaptive format means, and whether it changes how students should approach the test. Some of the AI answers are genuinely solid. Some are mealy-mouthed mush. And at least one is just flatly wrong in a way that could cost a student a shot at their target school.
We asked AI about something that derails high-potential students more than almost anything else: anxiety, perfectionism, and the gap between how a student performs in practice and how they perform when it counts.
We took a set of questions about performance under pressure: careless mistakes, running out of time, the difference between knowing material and being able to use it. Then we had Wes respond.
This time, we asked about something that stops a lot of strong students cold: the plateau. Why scores stall in the 1450–1500 range. Why more practice doesn’t seem to help. And what’s actually going on.
The AI answers are organized, confident, and not obviously wrong. A few are even mostly right.
But “mostly right” and “right for your kid” are not the same thing. And with elite school admissions, the gap between those two things is where families quietly lose.
AI doesn't give bad answers. It gives answers that sound complete. And for questions like these, that's actually the problem. Because the answers it gives are clean, logical, and just shallow enough to be dangerous. They sound right. They feel helpful. And they...