The adoption of #ArtificialIntelligence and #Automation technologies in the UK is still being stunted by a lack of digital skills in businesses, according to new research.
The study found that the main reason behind 81% of businesses not implementing AI is a shortage of talent in their workforce for building automation processes.
This is of particular concern given that leading companies in Government, Legal, HR, Healthcare and Finance are increasingly adopting AI or automation in one form or another.
There is virtually no industry that is expected to remain untouched by these technologies in the next five to 10 years.
Re-thinking AI education
With 81% of UK enterprise organisations also intending on investing in more AI solutions, it’s clear that to improve adoption levels businesses need to fundamentally change the way they are training staff.
In other words, education and specialist training is key if businesses want to be able to fully reap the rewards that AI promises.
If we wish to leverage the increasingly sophisticated technology available, we need to ensure that staff are well-drilled in how to use it effectively, ethically and safely.
The practical implications of this requirement include a lack of in-depth data and automation-specific training in all roles across businesses, as well as in related courses within further education. Given that the vast majority of companies plan on implementing these technologies in the coming years, every employee should be trained to not only understand tools and their capabilities, but also become competent in using them.
While part of the reason behind this evident skills gap will have been the unprecedented uptake in AI over the past ten years or so, this finding should be sufficient to kick-start businesses into taking serious action.
Black box to glass house
Additionally, too many organisations currently deploy so-called ‘black box’ solutions that are understood only by the data scientists who built them, and are inaccessible to the individuals affected; both users and customers.
The best way forward is to move towards transparent symbolic technologies that address the skills gap by making the technology more accessible to employees without a degree in data science. Such “low code” systems are easier to introduce, and to train current or prospective employees to use.
Increasing ease-of-use also benefits collaboration: intuitive, transparent systems are more portable, not only within but across businesses and industries, as their use appeals to a broad range of individuals and skill sets.
Another crucial consequence of increased transparency is ensuring employees across sectors are more comfortable with transparent AI and its readily interpretable outcomes.
All industries can benefit from technologies that model automated decision-making based on human expertise instead of complex algorithms that only its data-driven scientists can explain. This philosophy is especially important in regulated industries, where there is a real need for technology that can provide auditable decisions to satisfy customers and regulators alike.
To use a different term, this kind of ‘explainability’ is non-negotiable to any company using AI. In fact, Forrester recently found that 45% of AI decision makers say trusting the AI system is either challenging or very challenging, while 60% of 5,000 executives in an IBM Institute of Business survey expressed concern about being able to explain how AI is using data and making decisions.
To compound this point, at Rainbird our survey of 1,000 senior decision-makers found that 88% of financial services firms consider lack of transparency the biggest hurdle in adopting AI.
Industry and education must work together
There is an obvious lack of data skills among professionals in the very industries that stand to reap the biggest rewards from AI.
While a large part of addressing this problem lies with educators, it is hard for them to keep up with the pace of change. It is equally necessary that those behind these new technologies ensure they are transparent.
By adopting an approach centred around subject matter-specific expertise, and using a technology which is designed to be understood, we can succeed in enabling a greater proportion of the professional population to truly embrace and thrive in this new AI age, as both authors and consumers of automation systems.
James Duez, CEO at AI-powered, automated decision making platform, Rainbird