News

Shallow AI/ML Talent Pool Gets Deeper, Thanks to Schools and Vendors

• By John K. Waters
• 08/24/2018

Demand for artificial intelligence and machine learning (AI/ML) skills is exploding. Between June 2015 and June 2018, postings on the Indeed job-search Web site for AI-related roles almost doubled. During the same period, the percentage of searches on the site using "AI" or "machine learning" increased by 182 percent.

The top-10 list of in-demand AI/ML roles noted in a recent Indeed blog post included top-ranked "machine learning engineer," followed by "data scientist," "computer vision engineer," "algorithm engineer," principal scientist," "computer scientist, "research engineer," "statistician," "director of analytics" and "data engineer." More than 40 percent of those machine learning engineer jobs listed on Indeed's Web site remained open after 60 days. (The average salary offered for that hard-to-find machine learning engineer was $134,449, according to Indeed.)

The explosive demand for AI/ML capabilities in a range of systems and applications has surfaced a serious, industry-wide skills gap. A short-term solution might be found in online "skills elevating" courses, such as Microsoft's Professional Program for Artificial Intelligence, and such online certification programs as Udacity's Artificial Intelligence nanodegree offering.

But there's also a long term to consider, Indeed points out.

"Encouraging people, especially kids, to learn more about computer science could help prepare individuals for this new part of the labor economy and ensure that the important work around AI gets done," the company said.

Educators have been addressing the need for student computer literacy for a long time now, but they're also rising to this recent, specific challenge with programs designed to equip students with some of the fundamental skills they'll need to pursue careers in AI/ML. One recent example: Ignite My Future in School, described as "a one-of-a-kind initiative transforming the way students across America learn through computational thinking, a foundational skill for 21st century success."

Launched one year ago by Discover Education and Tata Consultancy Services (TCS), the initial program included 10 early adopter school systems comprising more than 6,700 teachers and 194,000 students. The program's aims to "bridge the gap by developing relevant skills for future job opportunities in the new digital economy, including Artificial Intelligence, machine learning and the Internet of Things (IoT)."

Discovery Education is a provider of digital education content and professional development for K-12 classrooms. TCS is a global IT services, consulting. and business solutions company. The program's creators aim to use it to engage 20,000 teachers and one million U.S. students by 2021.

Last week, Discover and (TCS) announced a jointly developed program that will provide a group of middle school educators from Southeast Michigan with the opportunity to participate in Ignite My Future in School's Day of Discovery at Belleville High School. The event will include professional development training and interaction with curriculum experts to learn about the program, understand its purpose, and develop strategies for integrating these new resources into classroom instruction.

Of course, the vendors are waiting around for the school system to produce an AI/ML-savvy workforce. They're reaching into K12 to make sure budding AI/ML genius get the right start. One great example: Microsoft's annual Imagine Cup student developer competition.

Not your grandfather's science fair, or even your mom's hackathon, the Imagine Cup "brings together student developers worldwide to help resolve some of the world's toughest challenges," the Web site states.

The competition has been around for 16 years and involved an estimated 2 million students from more than190 countries. The competition statistics show a clear expectation among the students that AI/ML technologies are going to play a big part in resolving those challenges.

Around 40 percent of the total number of entries in this year's competition involved AI-based solutions. Among the 49 finalists, the percentage was even higher, involving such AI tech as as image recognition, speech recognition, machine learning, deep learning. And a bunch of students built their own ML models using open source frameworks such as TensorFlow, or Microsoft's AzureML service.

"This year's participants weren't just developing unique and potentially game-changing technology experiences, but they are also actively sharpening the skills that will boost their success as developers, technologists and entrepreneurs for the next generation," wrote Charlotte Yarkoni, corporate vice president in Microsoft's Growth and Ecosystems group, in a blog post.

One short-term strategy for bridging this skills gap that could gain some traction might be thought of as "low-code development for AI/ML." NNData, a Fairfax, VA-based software and services company, this week launched an online SaaS solution called Smart Data. Part of the company's flagship NNCompass data management solution, NNData is "an easy-to-use way to manage data along with use case-focused machine learning algorithms for anyone to use without having any training as a data scientist or programming background."

Something to think about while we're waiting for the next generation to step up.