Main image of article Amazon Wants to Teach You Machine Learning Through… Music?

“Machine learning” has rapidly become one of those buzzwords embraced by companies around the world. Even if they don’t fully understand what it means, executives think that machine learning will magically transform their operations and generate massive profits. That’s good news for technologists—provided they actually learn the technology’s fundamentals, of course.

Amazon wants to help with the learning aspect of things. At this year’s AWS re:Invent, the company is previewing the DeepComposer, a 32-key keyboard that’s designed to train you in machine learning fundamentals via the power of music.

No, seriously. “AWS DeepComposer is the world’s first musical keyboard powered by machine learning to enable developers of all skill levels to learn Generative AI while creating original music outputs,” reads Amazon’s ultra-helpful FAQ on the matter. “DeepComposer consists of a USB keyboard that connects to the developer’s computer, and the DeepComposer service, accessed through the AWS Management Console.” There are tutorials and training data included in the package.

“Generative AI,” the FAQ continues, “allows computers to learn the underlying pattern of a given problem and use this knowledge to generate new content from input (such as image, music, and text).” In other words, you’re going to play a really simple song like “Chopsticks,” and this machine-learning platform will use that seed to build a four-hour Wagner-style opera. Just kidding! Or are we?

Jokes aside, the idea that a machine-learning platform can generate lots of data based on relatively little input is a powerful one. Of course, Amazon isn’t totally altruistic in this endeavor; by serving as a training channel for up-and-coming technologists, the company obviously hopes that more people will turn to it for all of their machine learning and A.I. needs in future years. Those interested can sign up for the preview on a dedicated site.

This isn’t the first time that Amazon has plunged into machine-learning training, either. Late last year, it introduced AWS DeepRacer, a model racecar designed to teach developers the principles of reinforcement learning. And in 2017, it rolled out AWS DeepLens camera, meant to introduce the technology world to Amazon’s take on computer vision and deep learning.

Machine Learning Salaries

For those who master the fundamentals of machine learning, the jobs can prove quite lucrative. In September, the IEEE-USA Salary & Benefits Salary suggested that engineers with machine-learning knowledge make an annual average of $185,000. Earlier this year, meanwhile, Indeed pegged the average machine learning engineer salary at $146,085, and its job growth between 2015 and 2018 at 344 percent.

If you’re not interested in Amazon’s version of a machine-learning education, there are other channels. For example, OpenAI, the sorta-nonprofit foundation (yes, it’s as odd as it sounds), hosts what it calls “Gym,” a toolkit for developing and comparing reinforcement algorithms; it also has a set of models and tools, along with a very extensive tutorial in deep reinforcement learning.

Google likewise has a “crash course,” complete with 25 lessons and 40+ exercises, that’s a good introduction to machine learning concepts. Then there’s Hacker Noon and its interesting breakdown of machine learning and artificial intelligence.

Once you have a firmer grasp on the core concepts, you can turn to Bloomberg’s Foundations of Machine Learning, a free online course that teaches advanced concepts such as optimization and kernel methods. A lot of math is involved.

Whatever learning route you take, it’s clear that machine learning skills have an incredible value right now. Familiarizing yourself through this technology—whether via traditional lessons or a musical keyboard—can only help your career in tech.

By learning about the top machine learning engineer interview questions and answers, you can also advance your career by acing the interview once you're finally there.