Last September, Marcus, a 31-year-old resident of Springfield, Missouri, discovered a flaw with Apple's HomeKit, the iPhone maker's program for hooking up smart home gadgets in iOS. Marcus had outfitted his entire house with an array of HomeKit-approved lights, thermostats and door lock. The setup was all controllable with an iPad sitting in his living room that he could talk to using Apple's voice assistant, Siri. Everything was going great until one Friday morning in September, his neighbor walked up to his front door right in front of him and yelled, "Hey Siri, unlock the front door," and surprise, surprise: the door unlocked. Marcus was enjoying life with the smart lock, which was made by San Francisco-based startup August Home, but found the incident unsettling and took the device off his door.

Google Assistant already helps us organize our shopping lists and find out about things we're interested in, but it stops short at actually letting us buy items. With a new feature rolling out to Pixel phones, it looks like that's about to change. A new option for setting up payment methods has appeared in the Assistant settings for Pixel users, bringing the anticipated feature one step closer to reality. First discovered deep in the beta version of the Google App back in January, it seems as though Google is ready to pull the trigger on voice-powered payments, a major step in Google Assistant's fight against Alexa and Siri. Head into the Settings menu of the Google Assistant screen on your Pixel phone and you'll see the new option for Payments under the Google Account section.

That's because these chatty device are thought to be adding the ability to handle voice calls. Both devices need to continue to earn their place in your home, and according to the Wall Street Journal, adding the ability to make and receive calls is high up the list of forthcoming features. If it happens, it's going to happen this year. The big difference between Google and Amazon in this regard is that Google does have an advantage in the form of Google Voice. Voice already offers voice calls, text messaging, voicemail, and call forwarding features, so Google could just hook Voice up to Home devices. Amazon does not have an equivalent service to simply bolt on.

Lab A: Virtual, augmented & mixed reality Monday 15 – Friday 19 May 2017 Cambridge Junction, Cambridge UK - This lab will take a practical and theoretical approach, framed within a peer-to-peer environment. There will be taught sessions covering Unity software, HTC Vive, HoloLens, 360 video, use of sound and project workflow. This will be balanced with theoretical talks on virtual, augmented and mixed reality by artists and our industry partners. Lab B: Artificial intelligence Monday 12 – Friday 16 June 2017 Leverhulme Centre for the Future of Intelligence (University of Cambridge), Cambridge UK - Working with our partners, this lab will create a critical context for exploring AI from all perspectives, including the neighbouring areas of data culture and internet of things. There will be a mixture of taught sessions on how AI works, along with an AI'jungle' of devices such as Alexa, Siri, Cortana and WolframAlpha.

Baidu is furthering its push into artificial intelligence after it announced the acquisition of Raven Tech, a Chinese startup that developed an AI voice assistant platform. Baidu confirmed it has bought the startup's tech, product and staff of 60. The deal comes a month after Baidu hired noted AI expert Qi Lu, formerly with Microsoft, as its COO and Group President. Baidu didn't reveal how much it is paying for Raven Tech, which is an alumni of the Microsoft Venture Accelerator and Y Combinator and has raised $18 million from investors like DCM Ventures and Zhenfund. Raven Tech's Flow product was likened to a Chinese version of Siri, but it has failed to take off.

Longer term, who knows the fate of the human agent. There is certainly plenty of impetus for automation, but for now it can seem like a cool gimmick (see the Henn na Hotel). If used on any scale, AI has to be proven to definitively ease the customer journey and ultimately grow revenue.

In the voice assistant market, Apple has Siri, Google has Google Assistant, Microsoft has Cortana. There are also voice assistants that aren't natively integrated into phones or computers, such as SoundHound's Hound and Amazon's Alexa. We have also heard that Samsung is planning their own voice assistant called "Bixby". This is why it doesn't really come as a complete surprise to learn that Huawei could be working on their own voice assistant platform as well. This comes from a report from Bloomberg in which people familiar with the matter told the publication about Huawei's plans.

Huawei, the third largest smartphone manufacturer in the world, is reportedly working on its own digital voice-activated assistant. The company's AI assistant will soon be competing with Apple's Siri, Google Assistant and Amazon's Alexa. Huawei has formed a team with over 100 engineers in Shenzhen, China which is currently in early stages of development for the new digital smart assistant, according to one of Bloomberg. Huawei's digital assistant will be aimed at Chinese customers meaning it will communicate in Chinese languages. One of Bloomberg's sources claimed that Huawei will continue to work with Google and Amazon when it comes to providing other services to its phones internationally.

In recent years, there has been growing interest in the study of automated playlist generation — music recommender systems that focus on modeling preferences over song sequences rather than on individual songs in isolation. This paper addresses this problem by learning personalized models on the fly of both song and transition preferences, uniquely tailored to each user’s musical tastes. Playlist recommender systems typically include two main components: i) a preference-learning component, and ii) a planning component for selecting the next song in the playlist sequence. While there has been much work on the former, very little work has been devoted to the latter. This paper bridges this gap by focusing on the planning aspect of playlist generation within the context of DJ-MC, our playlist recommendation application. This paper also introduces a new variant of playlist recommendation, which incorporates the notion of diversity and novelty directly into the reward model. We empirically demonstrate that the proposed planning approach significantly improves performance compared to the DJ-MC baseline in two playlist recommendation settings, increasing the usability of the framework in real world settings.

There has recently been considerable interest in completing a low-rank matrix or tensor given only a small fraction (or few linear combinations) of its entries. Related approaches have found considerable success in the area of recommender systems, under machine learning. From a statistical estimation point of view, the gold standard is to have access to the joint probability distribution of all pertinent random variables, from which any desired optimal estimator can be readily derived. In practice high-dimensional joint distributions are very hard to estimate, and only estimates of low-dimensional projections may be available. We show that it is possible to identify higher-order joint PMFs from lower-order marginalized PMFs using coupled low-rank tensor factorization. Our approach features guaranteed identifiability when the full joint PMF is of low-enough rank, and effective approximation otherwise. We provide an algorithmic approach to compute the sought factors, and illustrate the merits of our approach using rating prediction as an example.