To help us navigate the world of AI, who better to speak to than David Reid, Associate Professor of Computer Science at Liverpool Hope University, an expert in spiking neural networks (SNNs), a branch of AI. This is what he had to say about the key skills and AI jobs related questions.   

Firstly, why is there such an acute AI skills shortage in the UK?
DR: when I started in the 80s more there were women in STEM than there are now, we’ve gone backwards. A lot has got to do with perception, role models and teachers to help get people into AI. It’s not taught in schools and teachers don’t have enough time to learn about AI. So we don’t teach or promote it enough at school level – basic stuff about AI should be on the curriculum.

Is the UK lagging behind other developed economies?
DR: the UK is actually one of leaders in AI, think of DeepMind and lots of ideas from Spinnaker projects in Manchester, the AI communities we have and SMEs developing AI algorithms and hardware. The problem is we never grow our nascent AI industries as much as we could. There is a massive skills gap, as the number of computer science and AI specialists is growing gradually but the need is far steeper, and the gap is getting bigger. It’s the same in the US. There is more of a cultural appreciation of AI in China and Japan than in western countries. There are a lot of great universities and SMEs with new ideas in the UK, the problem is AI will grow so rapidly we’re not going have the resources to fuel that growth.

What about the jobs being created by AI?
DR: AI is like the internet was 20/25 years ago, there are lots of new jobs in the industry and not just for creative researchers or architects. There will be a raft of other jobs for trainers, as AI systems require training and maintaining. Different skill sets are required, not just technical ones but how to interact with tech and demonstrate that tech to people.

How do your AI courses prepare students for the world of work?
DR: we try to make them as practical as possible and get students to solve real problems using AI. They must not only think outside the box but not be scared to fail; very often there isn’t a correct answer, so it’s about coming up with partial answers. It doesn’t matter if you solve it to an optimal level, close to that level is good enough. In Liverpool for example there are lots of SMEs, whereas big organisations require different skills. Often these companies know they should have AI but don’t know why and that’s where our graduates can take them through the process of how we do it and why.

What will be the most in demand skills over the next FIVE years?
DR: the two big skills needed will be 1) a passion about the subject and willingness to learn at a deep level, so being inquisitive and 2) adaptability. Not just one skill but a whole raft of stuff, as there are so many different areas and lots of different ways of doing things. For example, virtual and augmented reality is big right now but there might be another branch that takes off, so adaptability is key as new tech and techniques understand how they fit in. AI is so ubiquitous with so many different sets of skills, it’s not a case of replacing or not replacing jobs, it will become part of our jobs. Soft skills are so important not just for researchers, trainers and maintainers, so we get our students to speak in front of audiences about how AI system works. The ability to communicate to the layman and explain how AI works and its implications is vital.

Should organisations start upskilling their employees?
DR: reskilling to have a basic understanding of AI and how it impacts on society is incredibly important. Even though the people most affected by changes in AI are all under 18, those who have existing jobs must reskill. It’s scary, but you don’t have to be a mathematician to understand and interact with AI, you become a practitioner. Only researchers and architects need to understand how it works, most people don’t need that depth of knowledge. What is important is to have an understanding of how to use it and the problems AI can get into it, so you need to be aware of computational bias, how data can influence results, to know that AI isn’t infallible. Companies should start to demonstrate AI and introduce courses on why AI fails and what happens when it does. AI does not make decisions independently from the data it receives, what happens with most AI systems is that they have learnt to learn from data through a rough set of algorithms and how that data was presented to it. Learning about how important data is when you present it to AI system, how it consumes and uses data and gets it wrong is hugely significant.

Check out our latest digital & tech jobs