Over two months since Apple Music landed on Amazon Echo, it's now set for another smart speaker that isn't the HomePod. A listing for the music streaming service has been spotted inside the Google Home app for iOS by a Mac Rumors reader. Though it's currently not live, the update (Apple Music was previously slotted in the app's "limited availability" section) indicates that it could be imminent. Until December, the HomePod was the only smart speaker with Apple Music availability. That changed with its arrival on the Echo, and its Google Home debut would put it on all three leading smart speakers.

Badri is the Senior Vice President, Technology at Vonage - Video Engineering. As I speak with business leaders from around the world, I'm continually surprised by two important realities that seem to go unnoticed and that are poised to transform the way companies engage with their customers. First, while artificial intelligence (AI) remains a buzzword, many people are still unaware of how advanced algorithms have become. We're not talking about a collaborative filtering algorithm that predicts which Netflix shows you'll want to watch next. Today's algorithms are able to mimic human decision-making on tasks as complex as composing music and predicting what topics are of interest to your Congressional representatives.

Attention mechanisms have seen wide adoption in neural NLP models. In addition to improving predictive performance, these are often touted as affording transparency: models equipped with attention provide a distribution over attended-to input units, and this is often presented (at least implicitly) as communicating the relative importance of inputs. However, it is unclear what relationship exists between attention weights and model outputs. In this work, we perform extensive experiments across a variety of NLP tasks that aim to assess the degree to which attention weights provide meaningful `explanations' for predictions. We find that they largely do not. For example, learned attention weights are frequently uncorrelated with gradient-based measures of feature importance, and one can identify very different attention distributions that nonetheless yield equivalent predictions. Our findings show that standard attention modules do not provide meaningful explanations and should not be treated as though they do. Code for all experiments is available at https://github.com/successar/AttentionExplanation.

From Vincross, the folks who brought you the HEXA robot, is a new Kickstarter for MIND KIT, a modular robotics development platform that you can use to build that robot that you've always wanted but doesn't exist yet. "A robot that keeps hugging you and won't leave you alone" is either the best thing ever, or a horror movie. Worth the risk, I'd say. It's nice that the system is so modular, and also that it looks like there's a strong emphasis on manipulation, which you don't always get from robotics kits that tend to be primarily mobile bases and not much else out of the box. There's also some built-in functionality, like map-making that works with the mobile base and lidar module. Specs-wise, the brain module contains a quad-core 1.2 GHz processor with 2 GB of RAM and 8 GB of storage.

In this tutorial, you will learn how to colorize black and white images using OpenCV, Deep Learning, and Python. Image colorization is the process of taking an input grayscale (black and white) image and then producing an output colorized image that represents the semantic colors and tones of the input (for example, an ocean on a clear sunny day must be plausibly "blue" -- it can't be colored "hot pink" by the model). The novel approach we are going to use here today instead relies on deep learning. We will utilize a Convolutional Neural Network capable of colorizing black and white images with results that can even "fool" humans! To learn how to perform black and white image coloration with OpenCV, just keep reading!

I am a major sci fi fan. Well, at least I thought I was until I went to my first Star Trek convention in my 20s and realized that I was in the minority of people who did not speak Klingon or know episode numbers, titles or dates. Most recently, I have become inspired by Black Mirror, a show originally aired by the BBC and now offered on Netflix. The brainchild of Charlie Brooker, Black Mirror is the Twilight Zone for our times, giving us a glimpse as to how technology trajectories can be used to affect society in unintended ways in the coming decades. As Frederik Pohl used to say, 'A good science fiction story should be able to predict not the automobile but the traffic jam.' Metaphorically speaking, this show sure is predicting traffic jams.

An ancient Japanese temple has hired the help of a £700,000 (Y100 million) robot to deliver the teachings of Buddhism. The android, dubbed Kannon, is based on the traditional Buddhist deity of mercy and is designed to attract young people. The robo-deity provided its first sermon at Kyoto's Kodaiji temple, which opened in 1619, on Saturday. The android is a the result of a collaboration between the zen temple and Hiroshi Ishiguro, professor of intelligent robotics at Osaka University. Only Kannon's head, neck shoulders and hands are covered in a skin-like material made from silicone.

Microsoft has made mixed reality less awkward. That was the major takeaway from my time with HoloLens 2, its new $3,500 headset aimed not at you (or me), but at businesses. It's not only because the headset is so much more comfortable, (though that can take a lot of the credit). Microsoft has also doubled the mixed reality field of view, making the holographic overlay less like a floating reticle and more like a (virtual) display that's useful. At its MWC press event, Microsoft claimed it was the only holographic device that can display 47 pixels per degree of field of view, making 8-point fonts readable.

LAST week, the OpenAI research group announced it had created an artificial intelligence capable of generating hundreds of words of convincing text on almost any topic (see Fears of OpenAI's super-trolling artificial intelligence are overblown). But the group said it wouldn't be releasing the AI, because of its potential to be used as a fake news generator. Fear over the power of fake news is widespread. Damian Collins, who heads a committee of UK MPs looking into the matter, this week proclaimed that "democracy is at risk from the malicious and relentless targeting of citizens with …

We may not have robots like C-3PO from "Star Wars" or Rosie from "The Jetsons," but there's no denying that the machines we do have are getting smarter. But are business leaders taking advantage of their capabilities? It's easy to imagine adopting AI in some indefinite future, but the technology already has the capacity to improve your ROI -- provided you actually implement it. While AI is making waves across industries, the logistics space stands to benefit more than most. According to a report from DHL and IBM, "the network-based nature of [logistics] provides a natural framework for implementing and scaling AI." It's no surprise, then, that companies are already experimenting with AI implementation, achieving small wins that help set the stage for much bigger ones. Small wins are key because they help test theories, establish best practices, and achieve buy-in from key stakeholders.