The past couple of days, I read a couple of long articles about some issues surrounding machine learning:1
- The Dark Secret at the Heart of AI. “No one really knows how the most advanced algorithms do what they do. That could be a problem.” From MIT Technology Review.
- Alien Knowledge — When Machines Justify Knowledge. I like the original title better: “Our machines now have knowledge we’ll never understand.”
- Elon Musk’s Billion-Dollar Crusade to Stop the A.I. Apocalypse. “You’d think that anytime Musk, Stephen Hawking, and Bill Gates are all raising the same warning about A.I.–as all of them are–it would be a 10-alarm fire.”
Recommended.
Update. Before you assume I think machine learning is all doom and gloom, rest assured. I don’t.
- Can machine-learning improve cardiovascular risk prediction using routine clinical data? This is one example of where Betteridge’s law of headlines fails.
Image from @swiftonsecurity on Tumblr.
2 responses to “Machine Learning Links”
great links. read the alien knowledge in azeems links (https://www.getrevue.co/profile/azeem) but i’m working through the rest of your list. very cool stuff.
if our machine overlords explain what they’re up to, will we actually trust them?
Thanks for the pointer to Azeem Azhar: The Exponential View. It looks good. I read Dave Pell’s NextDraft. If you’re not reading it, I recommend adding it to your list.
Will we trust our machine overlords if they explain their reasoning? I think we’ll trust them, even if they don’t, until something goes wrong. Then there will be a huge overreaction. For example, imagine
ifwhen a self-driving car does something horribly wrong. When the public asks why and the engineers answer, “We don’t know why and the car can’t tell us,” how long until people start pressuring the government to ban self-driving cars?