I’m using my iPad mostly to take handwritten notes in graduate-level statistics classes, such as where we’re learning how to write our own machine learning algorithms. It’s great … the professor typically posts a PDF of lecture slides, and I can scribble directly on them, insert blank pages for the chalkboard transcripts, and so on. If I need to pull up something from a Web site or reference an email or check my calendar or whatever, it’s just a quick swipe away.
I can guarantee you that, as damned impressive as ChatGPT assuredly is, it is utterly useless for any sort of math more advanced than basic Calculus (and maybe not even that).
For example … I asked it to prove that the square root of two is irrational. It did well enough that it’d probably get an “A” on that question. Then, in the same chat, I asked it to prove that the square root of two is rational. It happily obliged with a convincingly-confident but seriously flawed “proof.” I then pointed out the most obvious contradiction between its first and second replies, and it stubbornly stuck to its guns that the square root of two is both rational and irrational.
I can’t wait for Siri to advance to the level of ChatGPT. And it’ll probably be sooner than any of us realize that it’ll gain graduate-level competency in various fields. But it’s still got quite a ways to go.
Right now, I’d say it’s basically a talented but overconfident ten-year-old with an encyclopedic photographic memory. Trust it accordingly.
b&