London Fashion Week and we’ve been on the hunt for the most innovative use of Machine Learning on the catwalk. Google and Zalando launched their ML 3D project and Tommy Hilfiger released a Facebook Chatbot while Lyst brought AR into the fashion world by using real, naked models as a kind of hybrid mannequin on which people can view clothes.
To kick off London Fashion Week, online retailer Lyst brought AR into the fashion world – by using naked models as a kind of hybrid mannequin on which people can view clothes. The models lined up semi-naked in a shop window in London on Friday morning and people outside the shop then held up phones or tablets to clothe the models digitally.
Fast food, along with shipping warehouses, has become a hot spot for robots and other automated technology. Dominos for example has numerous ways to place an order that does not involve talking with a human. The question if workers will be replaced by robots depends on whether it will make stores more efficient. Starbuck e.g. is making use of automation reduce the time spend on taking orders, but has not yet cut its workforce in any meaningful way.
“In our latest study, we estimate average fast food establishment will switch 1.2 workers from counter service to other tasks as remote order taking, delivery by robotic applications grow,”
In order to make chatbots more human-like, companies are creating teams of writers, poets, and novelists to write the responses we get from them. For example Cortana, Microsoft’s AI, was created with the help of 22 writers around the world.
James Canton, former Apple executive and global futurist, recently gave a TEDx talk on the future of thinking machines. His main idea was that we could be entering an era of a AI Dividend, where we use AI to solve challenges such as global hunger, disease prevention and climate change to generate a prosperity boost to the global economy.
According to Canton, we have to use AI to focus on solving the big complex problems of our era. He predicts that if we can solve these global challenges with the help of AI, over $100 Trillion US in prosperity could come.
Neural Nets overfit data and there are many techniques to regularise the Network towards reducing bias and variance. The author explains what Bayes’s Theorem is, and how you can use it to apply weight decay and weight constraints to reduce variance and improve generalisations.
GitHub user ppwwyyxx created a Neural Network toolbox on TensorFlow. You can use it to abstract your training task into three components: Model, Data and Callbacks. With the three components defined, Tensorpack trainer will run the training iterations for you.
Honeywell published their results of their Data’s Big Impact on Manufacturing: A Study of Executive Opinions, survey. More than 200 North American manufacturing executives took part in the study.
According to the results, 68% of manufacturers consider big data analytics and the IoT as essential for driving greater operations performance. In addition, equipment breakdowns and alleviate unscheduled downtime can be reduced by 26% and 23% respectively. The main two areas where big data analytics can have the greatest impact are avoiding unplanned downtime and improving supply chain performance.
This is the first instalments of a series which examines how the 4th industrial revolution will change our work and lifestyles. Driverless taxis and delivery services will be driven by nation’s shrinking populations as it is causing depopulation and labour shortages. Industrial revolutions of the past have replaced the work of humans but the 4th could make an even greater impact. According to Japan’s Trade and Industry Ministry, 7.35 million jobs, will be lost during fiscal 2030 if no measures are made to catch up to technological innovation.