The skills gap is the biggest barrier to AI adoption - but it doesn't have to be

If industry thought differently about what they want from candidates, the recruitment pool would open up, says Dr Patricia Charlton

Every part of the technology industry is facing the skills gap, but AI is suffering more than most. According to a 2017 study, there were just 300,000 professionals with AI skills worldwide two years ago, but ‘millions' of jobs.

Dr Patricia Charlton, Senior Lecturer at the Open University, told delegates at Computing's AI & Machine Learning Live event last week that the industry needs to stop focusing so much on qualifications.

"When you look at the [data science] job roles [being advertised] and you look at the people coming through, what you see is that a lot of the job roles ask for a number of things. Students have to have a minimum of an MSc, they have to have about two years of experience, and ideally a PhD. And of course that means to come through with that expertise is quite tricky."

Although AI faces issues like data challenges, company culture and hardware and other resources, the skills gap is the biggest barrier to adoption, said Charlton, who is also involved with the Institute of Coding (IoC).

The IoC is a collaboration between industry and academia, which is looking at ways to address the skills gap. This includes looking at key areas like graduate programmes, the apprenticeship model and diversity and inclusion. The IoC is

"[We're examining] the challenge of coming out as a graduate and deciding your career. So imagine you're starting your career today, you're thinking you're going to be a data scientist, or you're going to work in the tech industry - how do you make those decisions?"

She continued, "You might have seen a job [posting] today with a really long list of skills that you have to have - machine learning's often thrown in there, as well as two years' experience, a Masters programme.

"But we have many talented people coming through doing different kinds of degrees and qualifications. We have people who choose not to take a degree, who possibly would be very talented in those areas; people who might have had difficult or challenging periods in their lives, so they weren't able to complete their studies. So we're looking at a new way to think about that student employability journey."

The IoC is looking at a co-defined industry standard to address these issues. Charlton mentioned the SFIA Framework, which is a way of tracking and measuring skills and competency that is used widely in industry, but not in academia.

"[We are looking at] a relationship where industry understands what the graduate is coming out with, and academia understands what some of the industrial needs might be," she said.

"I don't believe, in three years, you can know all those many, many skills in great detail and just easily walk into a position...being work-ready. There's a transition; even if you take a new job today, [and] you've worked for many years, there's always a transition. We have to be realistic."