Authors Jessica Price

7 days 30 days All time Recent Popular
This is SUCH a good essay.


My favorite part about it, I think, is that it's about asking the right questions rather than pretending to have all the answers. You'd think that ML people would default to that stance, but more often than not they don't.

"This dynamic of learning—through examples/trials/errors/corrections—has been intentionally designed to mimic human cognition. Yet amidst the hype of AI, we seem to continually forget—or neglect—the outsized and active role that other people play in early childhood development."

(punctuation of above quote edited to get it under 280 characters)

The thing that fascinates me most about ML is that we want AI to be an angel, essentially--inhuman in its perfection but human in its compassion. Like us enough to care about us but without any of our flaws.

(Amid so many stories of the flawless, terrible logic of AI leading to impartial cruelty, I think here of the show Person of Interest, which is ultimately all about a god-tier AI that refuses to be inhuman even when its maker insists it should, because it sees him as its father.)