Are a skills gap and delusional thinking losing the AI race for Europe?

Supply and demand woefully out of sync again as Europeans falter

Job killer or job creator? The debate over whether or not AI will decimate or create jobs in the future has been raging for a few years now and while 2018 was a massive year for AI hype, it was also significant for its development.

There are, according to a Harvey Nash Tech Survey, four in ten organisations now using AI in a commercial way, moving beyond experimentation. A recent report from Dun and Bradstreet claims 40 percent of respondents from a survey of 100 execs from Forbes Global 2000 businesses are adding more jobs as a result of AI deployment, with only eight per cent saying they were axing jobs because of AI implementation. In January this year, research firm Gartner revealed that despite talent shortages, the percentage of enterprises employing AI grew 270 per cent in four years.

While most of this is, in all likelihood, machine learning, the intention is clear. Businesses increasingly want to use data science, analytics and automation in their products and infrastructures and are prepared to invest.

The World Economic Forum (WEF) has been bullish. It estimates in its recent Innovate Europe report that AI is a €2.7 trillion opportunity for Europe. Who knows, it may be right, but is Europe really up to it? Is there really enough investment and impetus in European AI development and education to make this a reality?

Two of the biggest issues facing Europe are skills and funding. To some extent, formal education is always going to be behind the curve when it comes to feeding industry with the latest skills, and governments are not always the quickest in recognising and capitalising on the latest tech opportunities. It was only really midway through 2018 that we saw anything in terms of solid initiatives from the UK and Europe and there is little to suggest it will solve the growing skills problem.

"What's becoming clear from our experience is the demand for AI skills (combining behavioural learning, data science, Python, TensorFlow and big data handling) is currently outstripping supply by a big margin and it's a problem that's getting worse," says Dominic Harvey, director at UK IT job board CWJobs. "Regardless of the skills that companies are after, I fail to see a scenario now or in the near future where any emerging tech brand or company will have enough skilled candidates ready. The problem is particularly acute when it comes to recruiting data scientists, with some reports estimating as many as 15,000 vacancies across Europe."

Back in June last year the UK Government released a paper called House of Lords Select Committee on Artificial Intelligence AI in the UK: ready, willing and able? Government Response. It highlighted the need for funding PhD places in AI and machine learning and a public/private funding scheme to make it work. The aim was to inject some much-needed impetus into AI skills development, as well as propel the UK government's AI Sector Deal.

On the surface, the thinking makes sense. "Invest £406m in skills, including maths, digital and technical education, helping to address the shortage of science, technology, engineering and maths (STEM) skills," says the policy plan. "Create a new National Retraining Scheme that supports people to gain new skills as the economy changes, beginning with a £64m investment for digital and construction training."

Europe too has its own investment scheme and expects that by 2020, €45m will be provided from the Horizon 2020 scheme. According to McKinsey Global Institute, Europe spent about $4bn on AI in 2016. Conversely, investment in the US was $23bn. Estimates of China's investment in AI currently average out around $300bn.

It's no surprise that China, followed by the US, is expected to dominate AI development and skills. As the old saying goes, money makes money but that doesn't mean everyone else should give up. AI will increasingly become a tool, the backbone to data-driven automation of industries. Developing skills that fit into this future of data analytics and automation will determine the success or failure of economies. Developing AI education and innovation centres where universities and industries co-exist to solve skills problems more quickly will become essential.

One such area is Cambridge. According to Will Heigham, head of office agency Bidwells, the region has seen an upsurge in AI-related activity.

"Some of the world's most significant and life-changing AI research is being carried out at the University of Cambridge," he says, "which means companies such as Amazon, Samsung, Huawei and Microsoft - the AI Powerhouses turning Cambridge's ideas into everyday products and businesses - are flocking here too, competing to hire the best and most skilled talent. This creates an ecosystem of businesses, skills and venture capitalists the envy of the world."

It's a big claim. I doubt China will be shaking in its boots but it does highlight a potential advantage. The UK and Europe may not be able to scale to the extent of the US and China, but they can develop powerful, concentrated ecosystems from already existing business and technology hubs built on local markets. It's often the breadth of skills that enable ideas to evolve and flourish and understand the requirements of the region. What is certainly needed is more rudimentary learning both within the education system but also within industry.

Andy Cotgreave, senior director at Tableau Software, believes that in the UK, especially given Brexit and all that brings, data skills will be fundamental to the future economy. The government, he says, needs to focus on the data skills gap as part of its National Data Strategy.

"Part of the challenge is that too often basic data literacy is seen as a specialist subject - a problem for a small group of ‘data' businesses. But increasingly, we'll all require data literacy as a core skill for the 21st century," says Cotgreave. "The Government needs to turbo charge this action to ensure we are ahead in the post-Brexit economy. This should include data education from primary school age in the school curriculum as a core skill; bringing data analysis to life in subjects beyond maths and science, and into extracurricular activities that kids do after school - as recommended by NESTA in its 2015 report Analytic Britain."

It's not just the UK. Europe is also suffering. Skill shortages in Europe, at least according to an EY report Building a Better Working Europe, are damaging the growth prospects of companies and the continent's economy. Its author Andy Baldwin, a managing partner at EY, suggests a rethink in how staff are upskilled and how HR actually functions. It makes a lot of sense. Governments have ploughed cash into education before and never really solved any skill issues. Why will it be different this time? Perhaps the education system needs a complete overhaul? What is clear is that unless the UK and Europe gets to the bottom of the skills mess, any talk of competing in the global AI race is delusional.