In my many years testing all sorts of gadgets, few things have elicited as much spontaneous joy as Anki's Cozmo, its adorable robot for kids. Mostly, that was because it had a personality. Behind all of the sensors, cameras and other hardware, there was a team of animators breathing life into it. Now, Anki is taking everything it learned from Cozmo and putting it in a bigger, more powerful home robot: Vector. And unlike Cozmo, you won't need a phone to play with it.

Avocado toast is the hottest lunch trend to hit my apartment in years, and Lenovo just kicked it to a new level. After sticking Lenovo's Smart Display in my kitchen this week, my snack of choice has only gotten better, and I couldn't be happier. With the screen's built-in recipe search and step-by-step guide to cooking, I learned that rubbing the bread with some cut garlic adds a lot of flavor, and finishing it off with a sprinkling of parsley, lime, and chopped tomatoes does wonders. I may even toss a fried egg on my green toast this week. I could have looked up all these avocado toast recipes on my Android phone, but I never did.

Twitter will not suspend Alex Jones or Infowars because they have not broken any rules, its CEO has said. Just about every other major technology company – including Apple, Facebook, YouTube and Spotify – have all acted to either completley ban or punish the controversial conspiracy theorist radio host. But Twitter says that his posts are not in contravention of any rules and so he will stay up. That is despite Jones having used the platform to amplify his beliefs, which include claims that the Sandy Hook school shooting did not actually happen at all and that those mourning murdered loved ones were paid actors. The I.F.O. is fuelled by eight electric engines, which is able to push the flying object to an estimated top speed of about 120mph.

Samsung said Wednesday it will spend the sum, amounting to 25 trillion won, to hire artificial intelligence researchers, to be a global player in next generation telecoms technology and to boost its presence in electronics components for future cars.

Researchers at the Massachusetts Institute of Technology (MIT) have developed what is likely a world first -- a "psychopathic" artificial intelligence (AI). Stay up to date regarding Samsung's upcoming Galaxy Note 9, to be unveiled August 9. The experiment is based on the 1921 Rorschach test, which identifies traits in humans deemed to be psychopathic based on their perception of inkblots, alongside what is known as thought disorders. Norman is an AI experiment born from the test and "extended exposure to the darkest corners of Reddit," according to MIT, in order to explore how datasets and bias can influence the behavior and decision-making capabilities of artificial intelligence. "When people talk about AI algorithms being biased and unfair, the culprit is often not the algorithm itself, but the biased data that was fed to it," the researchers say.

If the 2016 election made one thing clear, it's that people in this country are getting their news from very different sources. But with the growing popularity of smart speaker-delivered morning briefings, this may be about to change. As digital assistants becomes the go-to news source for more and more people, so too does NPR. And while there's nothing wrong with millennials rediscovering public radio, some are raising the alarm about media diet conformity. Reporters Without Borders, the Paris-based NGO that promotes freedom of information, has voiced concerns about voice assistants, suggesting they present a "threat to pluralist news and information."

The Yamaha MusicCast BAR 400 fits neatly under a TV while the wireless sub-woofer can easily be tucked away out of sight. It's possible to add on surround speakers wirelessly using MusicCast Surround speakers.Yamaha In the past year or so there have been a few attempts at trying to package surround sound for a flat screen TV alongside speakers for a music system. Unfortunately, until now sound bars and sound bases have indeed offered great sound to boost your TVs weedy speakers but they've not been able to offer wireless facilities like Apple's AirPlay or the ability to play hi-res music files. Well, that could be about to change with a brand new soundbar that Yamaha has announced today.

In automatic speech recognition (ASR) what a user says depends on the particular context she is in. Typically, this context is represented as a set of word n-grams. In this work, we present a novel, all-neural, end-to-end (E2E) ASR sys- tem that utilizes such context. Our approach, which we re- fer to as Contextual Listen, Attend and Spell (CLAS) jointly- optimizes the ASR components along with embeddings of the context n-grams. During inference, the CLAS system can be presented with context phrases which might contain out-of- vocabulary (OOV) terms not seen during training. We com- pare our proposed system to a more traditional contextualiza- tion approach, which performs shallow-fusion between inde- pendently trained LAS and contextual n-gram models during beam search. Across a number of tasks, we find that the pro- posed CLAS system outperforms the baseline method by as much as 68% relative WER, indicating the advantage of joint optimization over individually trained components. Index Terms: speech recognition, sequence-to-sequence models, listen attend and spell, LAS, attention, embedded speech recognition.

ABSTRACT This paper investigates the impact of word-based RNN language models (RNN-LMs) on the performance of end-to-end automatic speech recognition (ASR). In our prior work, we have proposed a multilevel LM, in which character-based and word-based RNN-LMs are combined in hybrid CTC/attention-based ASR. Although this multilevel approach achieves significant error reduction in the Wall Street Journal (WSJ) task, two different LMs need to be trained and used for decoding, which increase the computational cost and memory usage. In this paper, we further propose a novel wordbased RNN-LM, which allows us to decode with only the wordbased LM, where it provides look-ahead word probabilities to predict next characters instead of the character-based LM, leading competitive accuracy with less computation compared to the multilevel LM. We demonstrate the efficacy of the word-based RNN-LMs using a larger corpus, LibriSpeech, in addition to WSJ we used in the prior work. Furthermore, we show that the proposed model achieves 5.1 %WER for WSJ Eval'92 test set when the vocabulary size is increased, which is the best WER reported for end-to-end ASR systems on this benchmark. Index Terms-- End-to-end speech recognition, language modeling, decoding, connectionist temporal classification, attention decoder 1. INTRODUCTION Automatic speech recognition (ASR) is currently a mature set of widely-deployed technologies that enable successful user interface applications such as voice search [1]. However, current systems lean heavily on the scaffolding of complicated legacy architectures that grew up around traditional techniques, including hidden Markov models (HMMs), Gaussian mixture models (GMMs), hybrid HMM/deep neural network (DNN) systems, and sequence discriminative training methods [2].

The fault is in our human condition, not our stars. Power, politics, ethics and storytelling are part of Shakespeare's Julius Caesar. Around the Globe, tea from London to Kyoto inspires socio-cultural values, traditions, and connections that have nothing to do with tea ceremony and everything to do with ethics and forgotten values we need to consider in a world developing AI. It's not enough to seek the integration of robotics, cognitive systems and machine learning. Devoting a year to listening to global perspectives and occupations using AI around the world provided new insight in defense with an F-22 pilot, to storytelling Pixar Animation studios, to healthcare for children, material sciences, and university mechanical engineering, seeing connections across seemingly dissimilar domains.