People are starting to get to grips with the concept of artificial intelligence (AI) and how it can help retail: particularly online, where chatbots, search engines and product recommendation tools are already blazing a trail. But with up to ninety percent of sales still emanating from bricks-and-mortar stores and only a few practical use cases of AI in retail, Michael Goller, Chief Technology Officer at retail tech specialist, Detego, explains how machine learning can really make its mark in the fashion retail sector. We were one of the early adopters of machine learning in the development of our retail software and have launched a popular chatbot tool to help consumers get quick answers to simple stock enquiries without having to seek out a sales assistant or call customer service. Aside from a general foreboding that there's more talk about AI than action, and that machines might eventually outperform humans, we get the distinct impression that most retailers are still trying to understand the basics about what machine learning can actually do for their business and how AI will help with consumer engagement. So, here's our attempt to unravel the mysteries behind the topic of artificial intelligence in retail and see whether it's a passing fad, or something retailers should be paying more attention to.

We've got some simple tips you can take to keep Amazon's Alexa from eavesdropping into your conversations and potentially sharing them. Amazon shipping horror stories put the onus on the shopper to be careful when buying. For most Amazon shoppers, the experience is likely a seamless one: you search for a product, perhaps read reviews and see related items, and if you're an Amazon Prime member, you'll likely have it delivered quickly. It should be that simple – especially when Amazon's $178 billion in annual sales rank the company as the largest Internet retailer and marketplace in the world. Alas, shopping on Amazon doesn't always go smoothly.

Artificial Intelligence and Machine Learning ceases to be a miracle of science fiction and gradually becomes a part of our daily life. However, lots of us still do not fully understand what Machine Learning is and why it is so important. Machine Learning is a process in which the system processes a large number of homogeneous information, identifies patterns and uses them to predict the characteristics of new data. It is the idea of the existence of common algorithms that can tell us something interesting about the data set without having to write a code specific for the problem. Instead of writing the code, you pass the data to a common algorithm, and it builds its own logic based on them.

In our November 9th Retail Expert Series webinar, I spoke with Microsoft's resident bot expert and Director of Retail Solutions, Shish Shridhar, about the meteoric rise of chatbots and conversational commerce in the retail industry. We discussed why chatbots are so rapidly gaining traction with retailers and their most common use cases in the industry. Shish also shared his experiences on how to overcome common challenges, identify the right use cases, and get started. As a follow-up, I'd like to address a common concern, re-share our industry examples, and dive a bit deeper into some of the tools and capabilities offered by the Microsoft Bot Framework. "Is chatbot technology really ready?"

We revisit the importance of the individual units in Convolutional Neural Networks (CNNs) for visual recognition. By conducting unit ablation experiments on CNNs trained on large scale image datasets, we demonstrate that, though ablating any individual unit does not hurt overall classification accuracy, it does lead to significant damage on the accuracy of specific classes. This result shows that an individual unit is specialized to encode information relevant to a subset of classes. We compute the correlation between the accuracy drop under unit ablation and various attributes of an individual unit such as class selectivity and weight L1 norm. We confirm that unit attributes such as class selectivity are a poor predictor for impact on overall accuracy as found previously in recent work \cite{morcos2018importance}. However, our results show that class selectivity along with other attributes are good predictors of the importance of one unit to individual classes. We evaluate the impact of random rotation, batch normalization, and dropout to the importance of units to specific classes. Our results show that units with high selectivity play an important role in network classification power at the individual class level. Understanding and interpreting the behavior of these units is necessary and meaningful.

Amazon.com Inc. (NASDAQ: AMZN) saw its shares rise marginally on Wednesday after the company announced a new product in its Echo line. The e-commerce giant is expanding its virtual assistant to help users manage their closets and image better. Ultimately, Alexa will help users look their best with the Echo Look. Alexa on Echo Look is a style assistant that gives personalized recommendations, helps keep users' closets organized and provides fashion inspiration. Echo Look offers features like Style Check, Collections and suggestions of new pieces to complement users' existing wardrobe.

In this special guest feature, Sarah Kampman, VP of Product at Square Root, discusses why retail is positioned to reap the biggest benefits of data analytics today. Sarah shares why machine learning and AI may be the secret weapon to solving for the the challenge of operating hundreds -- even thousands -- of disparate store locations, and how the approaches can be used to drive store performance, increase alignment and impact decision making at every level of a retail org. As the VP of Product at Square Root, Sarah finds solutions for customers that they didn't even realize they needed. With 15 years in product management and more than 20 years in technology overall, Sarah specializes in creating long-term focus groups with valued clients, helping them use technology to meet their business needs. She understands that building relationships through empathy leads to the most dynamic ideas and strategies.

Adobe has released it Digital Trends report for 2018, specifically looking into the changing landscape of online retail, as well as the way digital users are reshaping the space. Despite predictions of doom and gloom for bricks-and-mortar, the strength of online sales doesn't mean that physical stores are obsolete. In fact, a 2017 US Census report found that 90 per cent of all retail purchases in America were made in physical stores. Furthermore, 43 per cent of Millennials prefer to research brands and products online before shopping in-store. This reinforces how critical it is that retailers deliver an exceptional and joined-up cohesive experience along the entire shopping journey.

Amazon now has more than 100,000 robots inside its warehouses worldwide. But despite the sheer volume of robotic employees working in its fulfilment centres, Amazon insists robots will not replace human labour entirely - at least, not yet. The firm claims the robots do not have the'common sense' or'dexterity' needed for the job. Amazon uses its robots to carry stock around the expansive warehouse floors and group together all the individual items needed for a specific order. This is done to reduce the amount of interaction humans have with the products.

Amazon's fleet of automated warehouse robots, now more than 100,000 machines strong, is working alongside human employees to help meet the e-commerce giant's massive fulfillment demand. The company's robots carry inventory around massive warehouse floors, compiling all the items for a customer's order and reducing the need for human interaction with the products. But the chief technologist of Amazon Robotics, Tye Brady, insists that these robots are enhancing human efficiencies rather than eliminating warehouse jobs. Amazon has been going full steam ahead when it comes to hiring and now employs over 500,000 people. Brady views the robots as necessary to this growth. "When there are tens of thousands of orders going on simultaneously, you are getting beyond what a human can do," he told the audience at MIT Technology Review's first EmTech Next conference today.