Copy
Machine Learnings

Awesome, not awesome.

#Awesome
"There’s a broad spectrum of [story] interactivity, says Gagliano: “A lot of interactive films have decision moments, when you can branch the narrative, but I wanted to create something that let you transform the story at any point.” A certain degree of interactivity comes from choosing the type of AI that controls each character. You can make some use rule-based AI, which guides the character using simple heuristics—if this happens, then do that. Then you can make others become reinforcement-learning agents trained to seek rewards however they like, such as fighting for a bite of the fruit. Characters that follow rules stick closer to Gagliano’s direction; RL agents inject some chaos." - Will Douglas Heaven, Editor Learn More from MIT Technology Review >

#Not Awesome
""In many jurisdictions, facial recognition systems adopted at the border are in the process of being repurposed to achieve many unrelated public and private sector objectives," reads the report. It notes that the technology's "ability to identify otherwise anonymous individuals and pervasively link them to rich digital profiles" without individuals' knowledge or consent poses serious risks to human rights. " - Nathan Munn, Writer Learn More from Vice >

What we're reading.

1/ Tristan Harris, the start of The Social Dilemma on Netflix, responds to critiques on his films' attempts to show how social media platforms "inflame tensions" to keep us spending more time on them. Learn More from OneZero >

2/ Russia's efforts to install CCTVs with facial recognition software at the entrances to apartments and airports is a massive invasion of privacy. Learn More from Human Rights Watch >

3/ The head nurses at Duke Health hosted pizza parties for healthcare professionals to build the excitement needed to get Sepsis Watch, a deep-learning tool designed to help doctors prevent deaths, accepted in their hospitals. Learn More from MIT Technology Review >

4/ The award for best research paper at the 2020 Machine Learning for Signal Processing conference was given to a paper that can identify "what kinds of anomalies current methods can detect well and what kinds they might miss" in large datasets used to train machine learning models. Learn More from Yale News >

5/ Researchers at Harvard Medical school develop a new machine learning model that may help to predict dementia in older patients. Learn More from The Harvard Crimson >

6/ Many new recruiting tools allege the ability to help remove bias from the hiring process, but companies are starting to learn the hard way that the opposite effect is often true. Learn More from Quartz >

7/ Black box algorithms incentivize gig-economy drivers to break laws in order to deliver goods more efficiently. Learn More from Emerging Tech Brew >

Links from the community.

"Deep Drone Acrobatics (RSS 2020)" submitted by Samiur Rahman (@samiur1204). Learn More from YouTube >

"Algorithm discovers how six simple molecules could evolve into life's building blocks" submitted by Avi Eisenberger (@aeisenberger). Learn More from Chemistry World >

See these stories in Journal >

First time reading Machine Learnings? Subscribe to get the next edition >
11/10/20 View this email in your browser
Copyright © 2020 Machine Learnings, All rights reserved.


Done learning about AI?
Opt out of emails