Tao is worried that AI will produce drugs that hit metrics but don’t actually cure the underlying disease.
Does that sound familiar? That’s (most of) modern medicine!
Think about optimizing for 25(OH)D and LDL metrics.
That’s how you get people taking Vitamin D supplements and
This is a slide from Terence Tao's 'Math 2.0' public lecture yesterday evening at Caltech. I am highly confident that, if the cure actually worked, almost any Stage 4 cancer patient would tell you that human comprehension of its mechanism is completely irrelevant to them.
