4 ms·
He is also missing the point, perhaps intentionally. Of course we can understand how the learning algorithms work, since these are code written by us. But not h
by cubefox 3y ago
He is also missing the point, perhaps intentionally. Of course we can understand how the learning algorithms work, since these are code written by us. But not how the models work, since these are written by the learning algorithms, as they churn through the vast training data. At billions of parameters, models are black boxes.
- bamfly 3y agoSuch a weird direction to take it, total non-sequitur. It's like responding to concern about risks of nuclear war or bio-engineered pathogens with "pft, you can understand that tech, it's all in textbooks, it's just physics and engineering". Like... OK? So? That has nothing to do with it.