In the News
How A.I. can keep accelerating after Moore’s law
New ideas in chip design look likely to keep software getting smarter.
Customer service bots can detect your agitation
A virtual assistant that can tell you’re frustrated can slow down and help you out.
Sponsor
Test your developer skills and win Udemy course(s) worth $50 - Exclusive to the AIweekly community!
The new Developer Economics survey is a sci-fi adventure, so it is actually fun, plus you will learn about new tools! Talk about your dev habits related to your work and hobbies, whether it’s for your machine learning projects, desktop app development or any other. Start here!
*to be eligible for the prize, you need to complete the survey and submit your email
Learning
Training AI bots to negotiate
Interesting work by Facebook AI Research. They have built a framework to train AI bots on human negotiations. The bots seem to be learning human-like negotiation strategies on their own.
Convolutional methods for text
What can NLP learn from the popularity of Convolutional Neural Networks commonly used in Computer Vision? Great run down of convolutional methods applied to NLP.
If your company isn’t good at analytics, it’s not ready for AI
Companies that rush into sophisticated artificial intelligence before reaching a critical mass of automated processes and structured analytics can end up paralyzed.
The machine intelligence behind Gboard
Neural spatial models and finite-state transducers power Google's "intelligent keyboard".
Software tools & code
Extreme event forecasting at Uber with RNNs
At Uber, event forecasting enables us to future-proof our services based on anticipated user demand. The goal is to accurately predict where, when, and how many ride requests Uber will receive at any given time.
MobileNets
Open-source models trained and designed to be lightweight and faster. They are less accurate than "standard" ResNet and Inception networks but a lot faster and lighter for on-device analysis.
Exploring LSTMs
LSTMs are a fairly simple extension to neural networks, and they're behind a lot of the amazing achievements deep learning has made in the past few years.
Some thoughts
Human-level A.I. is right around the corner—or hundreds of years away
Ray Kurzweil, Rodney Brooks, and others weigh in on the future of artificial intelligence
Training Imagenet in 1 hour
While training Imagenet in 2012 took a couple of weeks with a GPU and lots of custom programming, Facebook claims to have trained Imagenet in just 1 hour recently, using 256 GPUs.
About
This newsletter is a collection of AI news and resources curated by @dlissmyr. If you find it worthwhile, please forward to your friends and colleagues, or share on your favorite network!
Share on Twitter · Share on Linkedin · Share on Google+
Suggestions or comments are more than welcome, just reply to this email.
Thanks!