High-Stakes AI Decisions Need to Be Automatically Audited
Opinion: The current standard for evaluating AI is insufficient. AI systems should be instantly interrogated for bias by a third party.
The Toxic Potential of YouTube’s Feedback Loop
Opinion: I worked on AI for YouTube’s “recommended for you” feature. We underestimated how the algorithms could go terribly wrong.
This New Poker Bot Can Beat Multiple Pros—at Once
Software designed by a Carnegie Mellon professor and Facebook researcher beat five human pros in No Limit Texas Hold ‘Em.
Who’s Listening When You Talk to Your Google Assistant?
A Belgian broadcaster obtained recordings of more than 1,000 conversations with Google’s automated assistant, some including personal details and medical information.
Taser Maker Says It Won’t Use Facial Recognition in Bodycams
Axon, the maker of Tasers and police bodycams, said it won’t deploy facial recognition systems, after a company ethics board recommended against it.
YouTube Is Giving You More Control Over Video Recommendations
You can now block certain channels from showing up in your “Up Next” list.
A Device to Detect ‘Aggression’ in Schools Often Misfires
Screams by high schoolers didn’t trigger the detector, but some coughs did. So did cheers for pizza.
Artificial Intelligence Is Coming for Our Faces
Trained for a week on a massive data set of portraits, a neural network spits out striking images of nonexistent people.
Deepfakes Are Getting Better. But They’re Still Easy to Spot
New research shows how to create seemingly realistic photos or video from a single image. But it won’t be easy to deploy anytime soon.
Some US Cities Are Moving Into Real-Time Facial Surveillance
Chinese authorities apply facial-recognition programs to surveillance videos to identify subjects. Now, a report says Chicago and Detroit have purchased their own systems.
Social