[dreamcompiler]

This completely explains why so many engineers are skeptical of AI while so many managers embrace it: The engineers are the ones who understand it.

(BTW, if you're an engineer who thinks you don't understand AI or are not qualified to work on it, think again. It's just linear algebra, and linear algebra is not that hard. Once you spend a day studying it, you'll think "Is that all there is to it?" The only difficult part of AI is learning PyTorch, since all the AI papers are written in terms of Python nowadays instead of -- you know -- math.)

I've been building neural net systems since the late 1980s. And yes they work and they do useful things when you have modern amounts of compute available, but they are not the second coming of $DEITY.

[LTL_FTC]

Linear algebra cannot be learned in a day. Maybe multiplying matrices when the dimensions allow but there is far more to linear algebra than knowing how to multiply matrices. Knowing when and why is far more interesting. Knowing how to decompose them. Knowing what a non-singular matrix is and why it’s special and so on. Once you know what’s found in a basic lower devision linear algebra class, one can move it linear programming and learn about cost functions and optimization or numerical analysis. PyTorch is just a calculator. If I handed someone a Ti-84 they wouldn’t magically know how to bust out statistics on it…

[tolciho]

> This completely explains why so many engineers are skeptical of AI while so many managers embrace it: The engineers are the ones who understand it.

Curiously some Feynman chap reported that several NASA engineers put the chance of the Challenger going kablooie—an untechnical term for rapid unscheduled deconstruction, which the Challenger had then just recently exhibited—at 1 in 200, or so, while the manager said, after some prevarications—"weaseled" is Feynman's term—that the chance was 1 in 100,000 with 100% confidence.