The 2020 LinkedIn U.S. Emerging Jobs Report identified the top 15 jobs over the previous five years and emphasized that “artificial intelligence and data science roles continue to proliferate across nearly every industry.” Artificial intelligence specialist (No. 1) showed 74% annual growth, and data scientist (No. 3) and data engineer (No. 8) followed with 37% and 33% annual growth.
But the problem is: we don’t have enough skilled talent to fill these jobs, and it’s a national imperative that we change the way we imagine, educate, recruit and upskill our technical workforce. We must abandon the flawed idea that AI jobs are only for people with master’s degrees or PhDs with decades of experience. We need to identify and embrace a “blue-collar AI” workforce in the U.S. — and building one will require the cooperation of educational institutions, companies and government agencies.
It’ll be worth the investment. Research from McKinsey has indicated that 82% of companies who have AI and ML are benefiting financially from their investments. Business leaders see clear value in AI and want to expand its footprint within their companies, which is producing a massive surge in demand for AI talent. We can “mind the gap” and meet this surge by doing things differently, more tactically and with more players on our team filling key roles, rather than holding our breath for a fleet of perennial unicorns.
The Current AI Challenge
Every week, executives spend hours interviewing candidates who are ill-suited for roles their companies seek to fill. The mismatch happens often because non-technical workforce recruiters and hiring managers develop sub-optimal position descriptions and misidentify the needs of their teams. I’ve likened it to a baseball team signing nothing but pitchers or only a few right-handed outfielders when they need more specialized hitters, runners and fielders. Not every player needs to be a Gold Glove catcher or MLB All-Star pitcher. Success relies on identifying, recruiting, incentivizing and managing a mixed stable to truly win the ballgame. It is, understandably, a challenge to do this when the technology and skills are constantly changing.
In my roles on the U.S. Department of Commerce’s National Technical Information Service Advisory Board and the Certified Analytics Professional’s Analytics Certification Board, I talk to many other leaders facing similar challenges. The U.S. is in a global race for AI dominance. China in particular has poured billions of dollars into AI technologies and education, and we haven’t been able to keep pace. The U.S. is hardly keeping up, and many companies end up outsourcing technical talent, hiring foreign nationals with advanced degrees for open positions.
I value diversity within organizations, but I also work in government spaces where a lack of technically skilled citizens able to obtain proper security clearances poses an enormous long-term problem to both our national security and national prosperity. It makes simple sense to invest more strategically in scaling national educational programs that develop and curate domestic workforces for many echelons of education, experience and skill. We need an army of AI workers, and many of the more common ranks should be educated in novel, well-focused specialties via two-year and four-year programs. Companies will always need some with master’s and PhDs, but to scale and compete, we will need legions in the blue-collar AI workforce.
The Opportunities For Colleges And Universities
A big part of the talent shortage is the confusion around which competencies companies are looking for and how they will be utilized. What skills do we truly need to power artificial intelligence, machine learning and data science? What do all these terms really mean?
AI is made up of a matrix of overlapping methods and areas of practice, including natural language processing, machine learning, UI/UX and data engineering. They are implemented in a wide variety of business applications, such as cloud computing, cybersecurity, finance, supply chain management and call centers.
High-level AI jobs, like a machine learning engineer, may require a master’s degree or higher, but many others do not — and that’s where we need the majority of new hires. These jobs are perfect for associate’s degree graduates. While some have coined the term “new-collar” jobs, I feel that the term misses the mark and has neither been understood nor applied. Let’s keep it simple: we quickly intuit what blue-collar AI means, and the community college and undergraduate models that worked for us to scale the industrial workforce can once again meet the challenge of scaling a digital workforce.
I advise colleges and universities to work with companies to build the demand for these programs. It’s a win-win-win for schools, employers and graduates. Colleges become the primary source of in-demand talent in automation, data engineering, data labeling and data visualization. Employers hire candidates with associate’s degrees for a starting salary of around $60,000, instead of going offshore and spending the same amount for someone with an advanced degree, and having to upskill and do additional on-the-job training. And graduates gain immediate employment in high-growth industries, with great prospects for income growth and further education, while saving hundreds of thousands of dollars in student debt.
Our Greatest Hope
Our greatest hope is a blue-collar AI workforce that possesses practical skills that scale. Three categories of blue-collar AI workers that we should develop in the workplace and throughout our education pipeline include: data farmers, data miners and data welders.
• Data farmers cultivate data. When data isn’t available or very sparse, farmers use statistics, simulation and other mathematics to seed, cultivate and harvest data, enriching business requirements with real, surrogate or synthetic data. Farmers are the masters of getting things done with the data you have, not the data you wish you had.
• Data miners extract critical data for business applications. Miners analyze data, extract the critical ore and chisel the patterns that provide immediate insights or can be leveraged in future analytic fittings. Miners often provide ad-hoc analysis to support a business hunch or alert you to what’s happening throughout your business mines.
• Data welders connect data through operations, analytics and applications to drive business decisions. They align data to an action taken by customers or businesses. The data is inextricably tied to a repeatable process, which moves the data science from being an ad-hoc insight toward integrated business application.
By developing the national and organizational army of data farmers, miners and welders we will strengthen economies and find security now and well into the future. This is how we develop a sustainable ecosystem for being global leaders in AI innovation. These relevant, practical new career paths and digital skills will empower people, organizations and nations to move from what the world knows as legacy blue-collar work toward the future of techno blue-collar professionals: AI, IoT, cybersecurity, quantum computing and far beyond.