On Friday afternoon, adult children sat sandwiched between their parents in green chairs that filled the spacious Amazon Meeting Center. The auditorium buzzed with excitement like it was college orientation, except the children were Amazon employees and the parents wouldn’t be lugging furniture into a four-person dorm room afterward.
In a twist on the traditional “bring your kids to work day,” over 7,000 Amazon employees and parents swarmed the South Lake Union complex for a glimpse into the younger generation’s work at the tech giant. Amazon’s fourth annual “Bring Your Parents to Work” day welcomed parents from all 50 states and as far as India and China. Parents toured Amazon’s Spheres featuring winding walkways and thousands of plants woven into mesh walls, and engaged in more than 15 presentations to learn about what goes on at the Seattle-based company’s headquarters.
The challenge facing Amazon presenters was to explain complicated subjects to this general audience in an accessible way. In this instance, how do you explain to your mother machine learning, an analytics system that finds patterns in a large amount of data?
In a nod to former generations, Amazon technical product managers Chris Morrow and Pierre Brunelle entered a stage at the front of the auditorium to the theme song for the longtime game show “The Price is Right.”
“Whether they know it or not, many Amazonians work directly on machine learning, or are influenced by [it],” said Morrow. In machine learning, workers annotate images, videos or text that are then fed into software to improve products such as Amazon’s voice assistant Alexa.
The presenters asked attendees to pull out their phones to answer multiple-choice questions through the interactive learning tool Wooclap, in a demonstration of training data sets for machine learning. Morrow flashed large signs bearing images of an alarm clock and yield sign, prompting participants to classify the objects by choosing the corresponding shapes on their phone.
“Classification is a type of machine learning that sorts things into categories,” explained Morrow. A real-world example of machine learning would be sorting customer reviews for products featured on the Amazon retail website, Brunelle added. For instance, a machine-learning model can be trained to classify millions of reviews as negative, positive or neutral in a matter of seconds, which allows Amazon to “detect defects on products or opportunities for improvement,” he said.
Parents interviewed after the event remarked on the interactive nature of the lesson and its relevance to products they use every day. Steve and Sue Gralla from Bermuda Dunes, Calif., joined the machine-learning presentation with their daughter Debbie Alexander, who works as a program manager for the retail website. Steve Gralla said he walked away with a better understanding of the topic, adding that it should be expanded into a “take your sibling to work day.”