Invento CEO Balaji Vishwanathan (right) and an employee interact with Spod, next to MITRI, a humanoid developed by the firm. At an office in HSR Layout, a box-shaped robot, mounted with a tablet, moves along the office floor while avoiding objects. As it detects a human face, it stops to greet and introduce itself: "Greetings, I'm Spod. I'm here to help you shop." Spod is an artificial intelligence-enabled robotic shopping assistant that visitors to supermarkets may well see in near future.

Some of Amazon's warehouse workers don't have to choose between work and play with the introduction of company-created video games that turn tedious tasks into productive fun. A detailed report from The Washington Post describes how the company has installed screens at many of its warehouse workers' stations that allow employees to turn tasks like assembling orders and moving items into competitive games. Game titles include options like MissionRacer, Dragon Duel, and CastleCrafter and typically involve a productivity-based point system. The more tasks a worker completes, the more points or progress they make in the game. Amazon is using video games to help increase productivity and make working in its warehouses less tedious.

The wide and rapid adoption of deep learning by practitioners brought unintended consequences in many situations such as in the infamous case of Google Photos' racist image recognition algorithm; thus, necessitated the utilization of the quantified uncertainty for each prediction. There have been recent efforts towards quantifying uncertainty in conventional deep learning methods (e.g., dropout as Bayesian approximation); however, their optimal use in decision making is often overlooked and understudied. In this study, we propose a mixed-integer programming framework for classification with reject option (also known as selective classification), that investigates and combines model uncertainty and predictive mean to identify optimal classification and rejection regions. Our results indicate superior performance of our framework both in non-rejected accuracy and rejection quality on several publicly available datasets. Moreover, we extend our framework to cost-sensitive settings and show that our approach outperforms industry standard methods significantly for online fraud management in real-world settings.

Amazon might want to consider placing more of its delivery drivers in uniforms, based on a comment from one of its higher-profile customers. Ice T, the rapper who later became a TV star on "Law & Order: SVU," posted a tweet on Tuesday claiming he almost shot one of the tech giant's drivers who was "creeping up to my crib" the night before to complete a delivery. "Message To Amazon: Now that you have regular people making your home deliveries.. Maybe they should wear a Vest with AMAZON DELIVERY on it," said Ice T on Twitter. In a follow-up tweet, Ice T said he wasn't mad at the delivery person but noted it's not safe for the drivers to operate without some type of uniform. Message To Amazon: Now that you have regular people making your home deliveries.. Maybe they should wear a Vest with AMAZON DELIVERY on it..... Amazon customer service reached out to Ice T on Twitter, saying his comments would be escalated to the company's logistics team for review.

Walmart has opened a store in Levittown, N.Y. that is intended to showcase the power of artificial intelligence. The store, announced last week, is packed with video cameras, digital screens, and over 100 servers, making it appear more like a corporate data center than a discount retailer. All that machinery helps Walmart automatically track inventory so that it knows when toilet paper is running low or that milk needs restocking. The company's goal is to create "a glimpse into the future of retail," when computers rather than humans are expected to do a lot of retail's grunt work. Walmart's push into artificial intelligence highlights how retailers are increasingly adding data crunching to their brick and mortar stores.

By Tushar Kaushik IN YOUR CART AI-enabled shopping assistant Spod can suggest products based on customer's age, gender. At an office in HSR Layout, a boxshaped robot, mounted with a tablet, moves along the office floor while avoiding objects. As it detects a human face, it stops to greet and introduce itself: "Greetings, I'm Spod. I'm here to help you shop." Spod is an artificial intelligence-enabled robotic shopping assistant that visitors to supermarkets may well see in near future.

In this Oct. 31, 2018, file photo, a man, who declined to be identified, has his face painted to represent efforts to defeat facial recognition during a protest at Amazon headquarters over the company's facial recognition system, "Rekognition," in Seattle. San Francisco is on track to become the first U.S. city to ban the use of facial recognition by police and other city agencies. These days, with facial recognition technology, you've got a face that can launch a thousand applications, so to speak. Sure, you may love the ease of opening your phone just by facing it instead of tapping in a code. But how do you feel about having your mug scanned, identifying you as you drive across a bridge, when you board an airplane or to confirm you're not a stalker on your way into a Taylor Swift concert?

Ever wished there was a way to know if your favorite Ben and Jerry's ice cream flavor is currently available in a grocery store near you? Instacart's machine learning team has built tools to figure that out! Our marketplace's scale lets us build sophisticated prediction models. Our community of over 70,000 personal shoppers scans millions of items per day across 15,000 physical stores and delivers them to the customers. These stores belong to our grocery retail partners like Aldi, Costco, Krogers, Safeway, and Wegmans.

AMAZON: Has opened an AI-powered convenience store in Seattle. The premise of Amazon Go is simple: to eliminate everyone's least-favorite part of the shopping experience, checking out. With ceiling-mounted sensors and cameras backed by artificial intelligence, Amazon is able to track every interaction a customer has with a product. It knows exactly when a product is picked up or put back. Go works like a physical manifestation of Amazon's 1-Click checkout, where you "click" by taking an item off a shelf.

In this thesis, we leverage the neural copy mechanism and memory-augmented neural networks (MANNs) to address existing challenge of neural task-oriented dialogue learning. We show the effectiveness of our strategy by achieving good performance in multi-domain dialogue state tracking, retrieval-based dialogue systems, and generation-based dialogue systems. We first propose a transferable dialogue state generator (TRADE) that leverages its copy mechanism to get rid of dialogue ontology and share knowledge between domains. We also evaluate unseen domain dialogue state tracking and show that TRADE enables zero-shot dialogue state tracking and can adapt to new few-shot domains without forgetting the previous domains. Second, we utilize MANNs to improve retrieval-based dialogue learning. They are able to capture dialogue sequential dependencies and memorize long-term information. We also propose a recorded delexicalization copy strategy to replace real entity values with ordered entity types. Our models are shown to surpass other retrieval baselines, especially when the conversation has a large number of turns. Lastly, we tackle generation-based dialogue learning with two proposed models, the memory-to-sequence (Mem2Seq) and global-to-local memory pointer network (GLMP). Mem2Seq is the first model to combine multi-hop memory attention with the idea of the copy mechanism. GLMP further introduces the concept of response sketching and double pointers copying. We show that GLMP achieves the state-of-the-art performance on human evaluation.