Awesome, not awesome.
“…We know we lose 2m brain cells every minute [a blood clot is in the brain].” Yet the two therapies that can transform [stroke] outcomes — clot-busting drugs and an operation called a thrombectomy — are rarely used because, by the time a stroke is diagnosed and a surgical team assembled, too much of a patient’s brain has died. Viz.ai’s [machine learning] technology should improve outcomes by identifying urgent cases, alerting on-call specialists and sending them the scans directly.” — Learn More from The Economist >
“…Transgender [Uber] drivers across the country are finding their accounts either temporarily or permanently suspended due to an Uber security feature that requires drivers to take a selfie to verify their identity. If the photo doesn’t come back as a match to other photos on file, it will get flagged….The Uber security feature is called Real-Time ID Check. It was rolled out in September 2016 to “protect both riders and drivers.” Occasionally, drivers will be prompted to pull over and take a selfie. That photo is then compared to the drivers’ photo on file, using technology from Microsoft Cognitive Services. If the photo isn’t a match, the driver’s account is temporarily suspended while Uber “looks into the situation.” — Jaden Urbi, Producer Learn More from CNBC >
What we’re reading.
1/ First Amendment advocates press Facebook to allow researchers on their platform to shed light on the “opaque” algorithms that guide much of our public discourse. Learn More from The Washington Post >
2/ Expect the immediate impact of automation on jobs to more closely resemble small changes than outright destruction. Learn More from The New York Times >
3/ Despite advancements in machine learning, national GDPs and standards of living have remained mostly unchanged for years. Techo-pessimists say past innovations were more transformative than ML. Techno-optimists argue that people are reaping massive benefits that accounting standards fail to measure. Learn More from Foreign Press >
4/ New musical artists fear what the future looks like if fans only discover new music from the playlists that Apple and Spotify feed them. Learn More from The Outline >
5/ If corporate leadership and recruiting professionals think turning to algorithms will completely remove bias from the recruiting process, they are wrong. Learn More from Bloomberg >
6/ Researchers are startled by the shortcuts that AI programs take to solve complex tasks. “Teach a learning algorithm to fish, and it might just drain the lake.” Learn More from WIRED >
7/ Researchers build a model that powers a virtual basketball player’s ability to dribble a ball, and it’s wild to see in action. Learn More from DeepMotion >
Links from the community.
“The Defense Department has produced the first tools for catching deepfakes” by Avi Eisenberger (@aeisnberger). Learn More from MIT Technology Review >
Join 40,000 people who read Machine Learnings to understand how AI is shaping our world. Get our newsletter >