AI News of the Week: Experts predict when machines will take your job


Here’s When Machines Will Take Your Job, as Predicted by AI Gurus

Many top minds think that automation will cost humans their employment, with up to 47% of all jobs gone in the next 25 years. And chances are, this number could be even higher and the massive job loss will come earlier.

So when will your job become obsolete? Researchers at the University of Oxford surveyed the world’s best artificial intelligence experts to find out when exactly machines will be better at humans in various occupations.


Self-Normalizing Neural Networks

Success stories of Deep Learning with standard feed-forward neural networks (FNNs) are rare. FNNs that perform well are typically shallow and, therefore cannot exploit many levels of abstract representations. We introduce self-normalizing neural networks (SNNs) to enable high-level abstract representations. While batch normalization requires explicit normalization, neuron activations of SNNs automatically converge towards zero mean and unit variance.


Facebook Trains ImageNet in 1 Hour

Facebook published a paper today detailing how they are able to train nearly 1.3 million images in under an hour using 256 Tesla P100 GPUs that previously took days on a single system.

The team reduced the training time of a ResNet-50 deep learning model on ImageNet from 29 hours to one.