From the Alan Turing Institute to DeepMind, the UK boasts a rich history and exciting present in machine learning and artificial intelligence research and development, led by academia and industry.
According to recent research, AI is the largest commercial opportunity for Britain, projected to add £232 billion to the UK economy by 2030. SMEs and start-ups will play a significant role in grasping this. The segment highlighted by Theresa May during her speech at the World Economic Forum in Davos in January, where she told world leaders that the UK’s strong start-up scene will be instrumental in making the UK a world leader in ethical AI. However, start-ups today still need to overcome some significant challenges before reaching their full potential.
Barriers to AI development
The next generation of ‘deep learning’ approaches have been reaching superhuman levels of performance recently. These achievements depend on having access to huge amounts of high quality training data, computation power and expertise in this area. DeepMind’s AlphaZero for example, relied on computational power greater than the World’s top ten supercomputers combined, to beat AlphaGo, the machine-learning software that conquered 18-time human Go champion, Lee Sedol, last year.
Access to this computation power and expertise is currently a significant barrier for SMEs. This was exposed by Digital Catapult’s recent report, Machines for Machine Intelligence, which found that the demand and cost of computation power for machine learning models is rapidly increasing. Individual AI and machine learning runs can cost upwards of £10,000, meaning that many AI start-ups can struggle to drum up sufficient investment to produce a proof of concept.
Finding people with the right skillset for AI development can be another challenge. According to one job site, the demand for software developers and machine learning engineers who create AI software has increased by 485 per cent since 2014. This leaves large and small organisations fighting tooth and nail over a shallow pool of talent – and potentially unable to meet their ambitions as a result.
Organisations researching and developing AI solutions are also being increasingly challenged on the ethics of their work. Several high-profile projects demonstrating that AI can propagate human biases have raised questions over the technology and the rules governing its development. It’s paramount that SMEs are aware of these, and the potential implications – to business and society – of inadvertently building bias into a product.
Finding the way forward
These barriers are substantial – and supportive initiatives will be paramount in helping promising UK start-ups overcome them and reach their full potential. Digital Catapult’s Machine Intelligence Garage is one such programme. Launched in December, the Garage provides access to computational power and expertise through partners including, Google Cloud, AWS and Graphcore. With access to these resources, companies will be able to bring their products to market faster, create better machine learning tools and scale-up their companies quicker. The programme also aims to educate start-ups about working with larger organisations, and large businesses about the benefits of AI adoption.
Strides are also being made in addressing the ethical issues around AI. The UK government has committed to opening a national Centre for Data Ethics and Innovation to ensure that ethical standards are at the forefront of AI and data-driven technology development. Companies’ consideration of data ethics has also been paramount when assessing applications for the Machine Intelligence Garage. The Programme, working alongside the National Centre for Data Ethics and Innovation, is building a world-class ethics process and group to help start-ups navigate these unclear and still unchartered waters.
With AI finding its way into more and more domains, the demand for tech talent is escalating. Education will be paramount in helping prepare future generations to fill AI vacancies. The UK government is already making strides in addressing this, with recent research indicating that initiatives to encourage young people into STEM careers are paying dividends, as individuals increasingly aspire to work in these professions.
Savvy organisations are also taking matters into their own hands. Some are looking to forge partnerships with universities to ensure a ready stream of AI talent, while others are forming internal learning centres to train employees in AI skills. Microsoft for example, has established an internal AI University, while computer chip developer, Nvidia, is providing teaching kits for educational institutions, covering introductory and advanced deep learning topics.
The UK has all the right ingredients to become a global leader in AI, however they need to be baked to perfection. This will depend on private and public-sector organisations coming together to tackle the challenges faced by AI businesses and those using AI. It is only through dialogue, collaboration and mutual support that these will be overcome, and the government’s ambition to make the UK a global leader in AI reached.
Marko Balabanovic is chief technology officer at Digital Catapult.