SEATTLE – A couple's private conversation was mysteriously recorded by their Amazon Echo device and sent to one of their contacts, igniting privacy concerns about the voice-activated gadgets the online retailer wants to make as commonplace in homes as televisions. The two found about it when they received a phone call two weeks ago from one of the husband's employees telling them to "Unplug your Alexa devices right now. You're being hacked," news station KIRO 7 reported. Inc. said in a statement to the station that it was an "extremely rare occurrence." The Portland, Oregon, couple used Amazon's voice-activated devices throughout their home to control heat, lights and the security system, according to the news report. Amazon in 2014 introduced the new line of devices, which can also stream music and order goods from Amazon via voice command.

Applying reinforcement learning in physical-world tasks is extremely challenging. It is commonly infeasible to sample a large number of trials, as required by current reinforcement learning methods, in a physical environment. This paper reports our project on using reinforcement learning for better commodity search in Taobao, one of the largest online retail platforms and meanwhile a physical environment with a high sampling cost. Instead of training reinforcement learning in Taobao directly, we present our approach: first we build Virtual Taobao, a simulator learned from historical customer behavior data through the proposed GAN-SD (GAN for Simulating Distributions) and MAIL (multi-agent adversarial imitation learning), and then we train policies in Virtual Taobao with no physical costs in which ANC (Action Norm Constraint) strategy is proposed to reduce over-fitting. In experiments, Virtual Taobao is trained from hundreds of millions of customers' records, and its properties are compared with the real environment. The results disclose that Virtual Taobao faithfully recovers important properties of the real environment. We also show that the policies trained in Virtual Taobao can have significantly superior online performance to the traditional supervised approaches. We hope our work could shed some light on reinforcement learning applications in complex physical environments.

The Home Shopping Network was a victim of its own success. Its reach had expanded across innumerable digital platforms--but while this generated reams of data about its customers, HSN's marketers struggled to keep track of individual retail journeys. To up its odds of driving conversions, the company needed to better understand its individual customers, from their unique tastes to their browsing habits. To make that possible, HSN knew it had to automate and accelerate data collection across its channels. It also had to deliver that data to marketers in an organized, easily understandable format that enabled them to take action quickly.

Ocado, the UK leader in home-delivered groceries from robot-run distribution centers, has established a licensing deal with US grocery chain Kroger (NYSE:KR) whereby Kroger will take a 5% stake in Ocado – an investment valued at $247.5 million and Ocado will help Kroger set up systems to manage online ordering, fulfillment and delivery operations utilizing Ocado-proven technologies. Ocado will see the Kroger chain build up to 20 Ocado-designed robot-run warehouses over its first three years. In a recent letter to Kroger stockholders, CEO Rodney McMullen said that Kroger is redeploying capital to emphasize improving its digital capabilities and enabling customers to shop in the store, by ordering online and picking up their order at the store, or getting their groceries delivered to their homes. Although McMullen didn't single out Amazon or any other competitors in the supermarket arena, Amazon's acquisition of Whole Foods and Walmart's price-cutting moves and partnering with online grocery deliver service Instacart are rapidly changing the landscape of grocery shopping. "Kroger is right in the middle of such a reinvention," McMullen said in the shareholder letter.

Amazon's facial recognition tool is being referred to as a'recipe for authoritarianism and disaster' after it was revealed to be used by law enforcement officials. Now experts say it raises even greater concerns, as the artificial intelligence used to power the technology could exhibit racial bias. Many are calling on Amazon to release data that shows they've trained the software to reduce bias, but it has yet to do so. A controversial facial recognition tool, dubbed Rekognition, marketed to police has been defended by its creator, online retailer Amazon. The controversy was spurred by a report from the American Civil Liberties Union (ACLU), which found that Amazon's facial recognition tool, dubbed'Rekognition', is being used by law enforcement agencies in Oregon and Florida.

Bots are dominating online communication, and most consumers don't even realize the number of times they have interacted with a bot while browsing their favorite website or completing an online purchase. Bot technology is far from a new concept, though, as its origins date all the way back to 1966 with the invention of Eliza, a bot which managed to mimic a natural conversation through programmed responses. With no sign of slowing down, here are five innovative ways bot technology is advancing. Artificial intelligence (AI) is a push towards being able to accomplish things that require human intelligence without needing humans. In the world of bots, the implementation of AI is geared toward creating environments where the bot you are speaking to is indistinguishable from a human being.

Never before have customers been more in control of the retail trade than today. Or has the retailer wrested control of the exchange? Let's revisit this in the light of new technologies and sensors deployed in this "game". In the sixties through the eighties, the Sears, Walmart and K-mart kind of super stores aggregated purchase information to decide what to buy and stock their shelves. Improving the scale of procurement they drove down the purchase price of things like Levi's Jeans to the detriment of the manufacturer.

Determinantal point processes (DPPs) have received significant attention in the recent years as an elegant model for a variety of machine learning tasks, due to their ability to elegantly model set diversity and item quality or popularity. Recent work has shown that DPPs can be effective models for product recommendation and basket completion tasks. We present an enhanced DPP model that is specialized for the task of basket completion, the multi-task DPP. We view the basket completion problem as a multi-class classification problem, and leverage ideas from tensor factorization and multi-class classification to design the multi-task DPP model. We evaluate our model on several real-world datasets, and find that the multi-task DPP provides significantly better predictive quality than a number of state-of-the-art models.

It's no secret that platforms like Instagram and Snapchat have transformed the way shoppers interact with brands. From influencers to sponsored posts, the way we shop has drastically changed in the last ten years. In response to this changing environment, many entrepreneurs have merged shopping with social platforms to create seamless experiences for both consumers and retailers. Chelfie, a Virginia-based social shopping app, is one of these solutions. Their platform allows users to receive live feedback on what to wear by sharing images privately or with a broader community of influencers.

Amazon founder and CEO Jeff Bezos laughs as he talks to the media while touring the new Amazon Spheres during the grand opening at Amazon's Seattle headquarters in Seattle, Washington, U.S., January 29, 2018. Amazon has been selling a facial-recognition system to police, sparking fears that the technology will one day power mass surveillance. On Tuesday, the American Civil Liberties Union and 35 other advocacy group sent a letter to the company's CEO Jeff Bezos, demanding that he stop providing the technology to law enforcement. The technology, called Amazon Rekognition, can identify people's faces in digital images and video. Police in Oregon and Florida have been using the system to help them solve crimes, but the ACLU argues that it's ripe for abuse.