Robots are taking over the world. Sure, you've heard that before. You even remember the Twilight Zone episode that warned us about it 60 years ago. One android recently published a novel. At Café X, in San Francisco, robot baristas make and serve coffee, and another California restaurant chain, Caliburger, is trying out a robot that can flip 2,000 burgers a day.

As the month comes to a close, we saw a bunch of solid tech deals across the web. Amazon's latest Echo Dot is down to just $30 and you can still save up to $200 on Samsung's Galaxy S21 smartphones. Fitbit's Mother's Day sale is in full swing, bringing record-low prices to many of its wearables, and May the 4th sales began early this year with discounted Star Wars Instant Pots. Here are the best tech deals we found this week that you can still get today. The latest Echo Dot and Echo Dot with Clock remain on sale for $30 and $40, respectively, which is very close to their all-time lows.

Research firm Gartner estimates the market for hyperautomation-enabling technologies will reach $596 billion in 2022, up nearly 24% from the $481.6 billion in 2020. Gartner is expecting significant growth for technology that enables organizations to rapidly identify, vet, and automate as many processes as possible and says it will become a "condition of survival" for enterprises. Hyperautomation-enabling technologies include robotic process automation (RPA), low-code application platforms (LCAP), AI, and virtual assistants. As organizations look for ways to automate the digitization and structuring of data and content, technologies that automate content ingestion, such as signature verification tools, optical character recognition, document ingestion, conversational AI, and natural language technology (NLT), will be in high demand. For example, these tools could be used to automate the process of digitizing and sorting paper records.

While Google Assistant more often than not gives us what we're asking for, there is still room for improvement in terms of accuracy. Google previously let us customize how sensitive we want the hotword to be on our smart displays and speakers. But now they're using federated learning with a new "Help improve Assistant" option to teach the system to avoid misactivations and misses. This lets you save audio recordings on your device to teach the speech technologies used for Google Assistant. As per 9 to 5 Google, this new option is turned off by default but when you go to your Google Assistant settings or your Google app's general preference, you'll see something called "Help improve Assistant" that you can turn on.

Amazon's latest top-of-the-range Alexa smart display has a trick up its sleeve like no other: it can follow you around a room. The third-generation Echo Show 10 costs £239.99 and is Amazon's largest smart display, sitting above the smaller £100 Echo Show 8 with an 8in screen and £80 Echo Show 5 with a 5.5in screen. It looks very different to its forerunners, like one of Amazon's Fire tablets mounted to a fat rotating Echo speaker, but fundamentally it is just like other wifi-connected smart displays powered by Alexa. The 10.1in screen is bright, good-looking and crisp enough for viewing at arm's length or more. It acts as a digital photo frame, displays timers and alarms, tells you the weather, the news and what's coming up on your calendar, controls smart home devices and plays music and videos from services such as Prime Video and Netflix, or YouTube through the built-in web browser.

In order for Google Assistant to really help users with daily tasks, the tool should be able to understand the human user. This means not only understanding the words you speak but also recognizing the meaning behind them. For this reason, the Assistant should adapt to your speech style without you needing to voice commands using specific words in the correct order. However, comprehending verbal speech proves challenging, given the highly variable context often used from one person to another. Yet another obstacle arises with words that sound different but are spelled the same way, such as two names.

We consider the problem of recommending relevant content to users of an internet platform in the form of lists of items, called slates. We introduce a variational Bayesian Recurrent Neural Net recommender system that acts on time series of interactions between the internet platform and the user, and which scales to real world industrial situations. The recommender system is tested both online on real users, and on an offline dataset collected from a Norwegian web-based marketplace, FINN.no, that is made public for research. This is one of the first publicly available datasets which includes all the slates that are presented to users as well as which items (if any) in the slates were clicked on. Such a data set allows us to move beyond the common assumption that implicitly assumes that users are considering all possible items at each interaction. Instead we build our likelihood using the items that are actually in the slate, and evaluate the strengths and weaknesses of both approaches theoretically and in experiments. We also introduce a hierarchical prior for the item parameters based on group memberships. Both item parameters and user preferences are learned probabilistically. Furthermore, we combine our model with bandit strategies to ensure learning, and introduce `in-slate Thompson Sampling' which makes use of the slates to maximise explorative opportunities. We show experimentally that explorative recommender strategies perform on par or above their greedy counterparts. Even without making use of exploration to learn more effectively, click rates increase simply because of improved diversity in the recommended slates.

Conversational Recommender Systems (CRSs) in E-commerce platforms aim to recommend items to users via multiple conversational interactions. Click-through rate (CTR) prediction models are commonly used for ranking candidate items. However, most CRSs are suffer from the problem of data scarcity and sparseness. To address this issue, we propose a novel knowledge-enhanced deep cross network (K-DCN), a two-step (pretrain and fine-tune) CTR prediction model to recommend items. We first construct a billion-scale conversation knowledge graph (CKG) from information about users, items and conversations, and then pretrain CKG by introducing knowledge graph embedding method and graph convolution network to encode semantic and structural information respectively.To make the CTR prediction model sensible of current state of users and the relationship between dialogues and items, we introduce user-state and dialogue-interaction representations based on pre-trained CKG and propose K-DCN.In K-DCN, we fuse the user-state representation, dialogue-interaction representation and other normal feature representations via deep cross network, which will give the rank of candidate items to be recommended.We experimentally prove that our proposal significantly outperforms baselines and show it's real application in Alime.

It was a first date out of a horror movie. Priscilla Castro, a 32-year-old from Vallejo, was headed to Vacaville on a Wednesday evening in August to meet Victor Serriteno, a 28-year-old she'd met through an online dating app. But instead of romance, the interlude ended in multiple deaths and hundreds of thousands of fire-scorched acres, prosecutors say. The person behind all the crimes, authorities say, is Serriteno, whom Solano County sheriff's deputies arrested again Wednesday on suspicion of arson and murder in connection with the Markley fire, which killed two people and merged into last year's devastating LNU Lightning Complex fire. Serriteno was already in the county jail after authorities alleged he killed Castro.

Most gadgets don't come with a user's manual that spells out every single feature. We learn them by doing, when someone spills the beans, or asking, "How'd you do that?" For example, no one thinks to dive into a new router's settings. The more connected devices you have, the more critical is this step. Tap or click here for a few essential steps to ensure your files, data, and network are safe from hackers and snoops.