Copy
Machine Learnings

Awesome, not awesome.

#Awesome
"In a new perspective paper, Stanford faculty discuss sex, gender and race bias in medicine and how these biases could be perpetuated by AI devices. The authors suggest several short- and long-term approaches to prevent AI-related bias, such as changing policies at medical funding agencies and scientific publications to ensure the data collected for studies are diverse, and incorporating more social, cultural and ethical awareness into university curricula." - Learn More from Stanford News >

#Not Awesome
"When Detroit police arrested Robert Williams with the help of facial recognition technology, they thought they may have found a suspect in a case involving the theft of $3,800 worth of watches. Williams says he was shocked – he'd never been in trouble with the law before. Facial recognition technology has proliferated among police departments across the country, who say it's been key in helping solve some of the toughest crimes. But with no well-established national guidelines governing how police should use the technology, three Black men have come forward since last summer alleging wrongful arrest due in part to facial recognition." - Learn More from CBS News >

What we're reading.

1/ AI systems can't just be dropped into an existing process without first considering the people that will be integrating them. Learn More from MIT Sloan >

2/ A team at intel uses machine learning to take existing graphics from GTA V, and upgrade them into a nearly photorealistic simulation. Learn More from The Verge >

3/ A Brooking's Institute analysis of 34 countries' plans for AI suggests that "democratic countries have signaled greater concern for ethics and governance of AI." Learn More from Brookings >

4/ A new machine learning system makes strides to better detect sarcasm, something that has been notoriously difficult to train AI systems to understand. Learn More from Defense One >

5/ A choose your own adventure game, powered by AI-generated text, took a deeply explicit turn that forced game developers to reconsider how to facilitate AI-human collaboration. Learn More from WIRED >

6/ One of the most potentially dangerous implications of advanced AI systems is human becoming too trusting of them (like people who fall asleep in their Tesla while it drives down the highway). Learn More from MIT Technology Review >

7/ How will humans adapt when an AI system develops better business judgement than the CEO? Learn More from The Guardian >

First time reading Machine Learnings? Subscribe to get the next edition >
16/05/21 View this email in your browser
Copyright © 2021 Machine Learnings, All rights reserved.


Done learning about AI?
Opt out of emails