Kevin Kelly: “The AI Cargo Cult: The Myth of a Superhuman AI”
Kevin Kelly often annoys me (see his What Technology Wants, but this is rather good, not least because it argues that the biggest flaw in the ‘superintelligence’-as-existential-risk argument is its assumption that intelligence is a one-dimensional attribute. Nobody who has read Howard Gardner can agree with that proposition.
“Jill Lepore on the Challenge of Explaining Things”
Terrific interview with the Harvard historian and New Yorker writer. Contains this wonderful passage:
“I only ever wanted to be a writer. I love history, and I especially love teaching history, but I never intended to become an academic, and I’m baffled by the idea that reaching a wider audience involves using smaller words, as if there’s some inverse correlation between the size of your audience and of your vocabulary. You don’t talk about, say, technological determinism to a freshman the same way you talk about it to a colleague, right? Is it easier to talk to a freshman? No, it’s harder. Is it more important to give that student a clear explanation of the concept than it is to chat with your colleague about it? I think so, though I suppose that’s debatable. I love the challenge of explaining things to other people, in the same way that I love other people explaining things to me. I love being a student. Nothing is so thrilling as diving into scholarship I’ve never encountered before and trying to get my bearings, learning what so many scholars have been piecing together over a very long period of time, and trying to figure out how to bring that learning to bear on a problem that I, like a lot of people both inside and outside the academy, happen to be struggling with. The hitch is getting the scholarship right. I always worry I’ve missed something, or distorted something, or failed to understand the big picture. That’s the downside: missing something crucial. Nothing is more concerning, or more discouraging, than getting something wrong; there’s no real way to right it. It’s horrible; it kills me.”
Maciej Cegłowski: “Build a Better Monster: Morality, Machine Learning, and Mass Surveillance”
Terrific, thoughful essay on the dystopia we have been building. ANd what we might do about it.