Get more from AI Weekly
Breaking stories as they happen. Deep dives on the topics you care about. 50+ free courses from Stanford, MIT, and more.
Deep Dives Daily Alerts Learning AIIn the News
Facebook AI
Deep-dive into Facebook's AI work and its quest to tackle the problem of emulating general intelligence — that is, getting computers to think less like linear, logical machines, and like us free-form humans — with a multi-prong approach. While the Facebook Artificial Intelligence Research (FAIR) team works on solving generalized AI problems, smaller groups like Language Technology and Facebook M deploy practical features to users.
Siri is savvier, but still not smarter than a 5th grader
Earlier this month, Apple announced that Siri was now "40% smarter". But is Siri really significantly better? This is a test comparing new Siri, old Siri and Google Now on a range of questions and IQ tests.
Humans with Amplified Intelligence could be more powerful than AI
"With much of our attention focused the rise of advanced artificial intelligence, few consider the potential for radically amplified human intelligence (IA). It’s an open question as to which will come first, but a technologically boosted brain could be just as powerful — and just as dangerous – as AI."
Robot swarms: scientists work to harness the power of the insect world
The hive mentality is inspiring the latest advances in technology and the US military is already experimenting with swarms of robotic boats and aircrafts
Also in the news this week...
- Google launches the beta of Cloud Dataproc, a managed service for Hadoop and Spark
- The Allen Institute claims it has developed an AI that scores as well as 11th graders on the SAT test
- Chatbot Rose wins the 2015 Loebner AI prize
- Deepmind claims its AI can now beat humans at 31 games, but still can't master Pac-Man
- Barbie will now come with Artificial Intelligence to chat
- Streamsets raises $12.5M to develop B2B big-data tool
- Clinicloud raises $5M to develop medical kit to diagnose illnesses with machine learning
Learning
Recurrent Model of Visual Attention
Recurrent attention models (RAM) are interesting ways of reducing the complexity of image analysis with ConvNets by adaptively selecting a sequence of regions or locations and only processing the selected regions at high resolution. This post discusses how RAM can be implemented, along with examples (code and test on MNIST).
Making meaningful restaurant recommendations at OpenTable
Slides shared at Recsys2015 on the approach OpenTable takes to restaurant recommendations (nearest-neighbor, collaborative filtering, word2vec...)
An Empirical Exploration of Recurrent Network Architectures
Recurrent Neural Networks are interesting but difficult to train. Some types of RNNs such as LSTMs are easier to train but are they optimal? This paper aims to explore thousands of possible RNNs to find optimal architectures for different types of tasks.
Software tools & code
Predicting cab booking cancellations
The business problem tackled here is trying to improve customer service for YourCabs, a cab company in Bangalore. The problem of interest is booking cancellations by the company due to unavailability of a car. The challenge is that cancellations can occur very close to the trip start time, thereby causing passengers inconvenience.
Password recovery with neural networks
Building a simple neural network that will recover password from a broken one, using a Discrete Hopfield Network
Neural Networks in JavaScript with Brain.js
Brain.js is a simple javascript library to create and train neural networks. The API has been simplified to a few methods calls and options, for integration in JS or node code.
Some thoughts
Is personalized discovery a feature, category or new paradigm
What do you do when you don’t know what you want to read, watch, listen to or do next? So far no one has created a solution that would automatically bring all the interesting options to your fingertips without you asking for it. A universal personalized Discovery solution doesn’t exist yet. Why?
About
This newsletter is a weekly collection of AI news and resources.
If you find this newsletter 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!
What do you think?
Join the conversation — share your take on this issue.
Log in to comment