It's been a big week for Google!
- They announced impressive revenue figures
- CEO Pichai said they are rethinking everything about Machine Learning.
- Google announced RankBrain, its new search AI
- Various Google scientists gave interviews, detailing how they would become highly dependent on AI in the future and what Google's AI challenges would be
In the News
Why self-driving cars must be programmed to kill
Self-driving cars are already cruising the streets. But before they can become widespread, carmakers must solve an impossible ethical dilemma of algorithmic morality.
Meet Google’s artificial intelligence chief
John Giannandrea is the company’s champion in Silicon Valley’s machine-learning arms race.
Inside the surprisingly sexist world of artificial intelligence
According to some women at the forefront of computer science, right now the real danger in the world of artificial intelligence isn’t the threat of robot overlords—it’s a startling lack of diversity.
Medical breakthrough in spinal cord injuries was made by a computer program
New software sifts through the information gathered in long forgotten studies and finds new avenues for researchers to pursue—like a new advance in treating spinal injuries. Great example of retroactive data mining.
Also in the news this week...
- Spain to invest 90M€ in natural language processing technologies
- Facebook unveiled its AI to describe photos for blind people
- Apple poaches top AI expert from NVIDIA to work on self-driving cars
- Law firm bosses said to envision Watson type computers replacing young lawyers
- Academic conference on "Love and sex with robots" canceled after being declared illegal
Workplace
Why data scientists need to be good data storytellers
Storytelling is data with a soul. Data Scientists are extremely good with numbers but numbers alone are not sufficient to convey the results to the end user. Being a good data storyteller is an art as well as a science.
Google Brain residency program
Google recently opened its Brain Residency Program. This is a 12-month role designed to jumpstart one's career in deep learning.
Learning
Techniques and systems for training large neural networks quickly (slides)
Presentation by Google's Jeff Dean on their work to quicken the training of large neural networks.
Estimating delivery times: a case study in Machine Learning
The problem in focus is predicting estimated delivery time for particular merchants. These help prospective customers get an idea of how long their orders might take.
Software tools & code
Squeeze the memory consumption of deep learning
One important theme about deep learning is to train deeper and larger nets, which is often a challenge as huge deepnet monsters are always hungry about the GPU RAMS. This article discusses how memory allocation optimization can be done for deep neural nets.
Brainstorm – Deep Learning library, successor to PyBrain
Brainstorm aims to make working with neural networks fast, flexible and fun. Its all written in Python.
Brains & Neurons
A fascinating aspect of language looks to be biologically hardwired in our brains
Does the Turkish word küçük (pronounced coo-chook) mean “big” or “small”? If you guessed the latter without knowing the language, you’re right—and there may be a cognitive explanation for your instinct.
Some thoughts
500 Deep Learning papers, graphviz and python
The author invested days creating a graph representing the evolutionary process of deep learning’s state of the art for the last 25 years. Impressive result.
What a Deep Neural Network thinks about your #selfie
Fun experiment to try to recognize good vs bad selfies with a dedicated deep neural network.
Visualize machine learning algorithms
Simple yet effective visualization of the results of different ML algorithms (js fiddle).
About
This newsletter is a weekly collection of AI news and resources. 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!