Effective government will lead to effective AI

What steps can policymakers take to help the UK make the most of machine learning and automation?

Sign Up

Get the New Statesman's Morning Call email.

In 2017, it was reported that more data was created in a single year than the sum of all information recorded by humanity in the past 5,000. Since then we have produced even more data year on year. In a world where social media sites claim as many followers as major global religions, and technologies such as the Internet of Things are shortly about to usher in an age of hyper-connectivity, this data explosion doesn’t show any signs of slowing down.

Exponential data growth means we will need an equivalent growth of people to process it and systems to help us derive value from it. In short, this means using artificial intelligence.

It may not be visible to consumers, but AI is starting to become mainstream inside organisations and is being used to tackle some of the challenges caused by this explosion of data. Analysis from the global market intelligence firm IDC shows that spending on AI systems reached $19.1bn at the end of 2018 and is predicted to hit $52bn by 2025, when 75 per cent of all organisations will be incorporating some aspect of AI into their operations. This means that organisations are already thinking about how to utilise AI, and governments and public sector organisations need to do so too. Where companies can use AI to drive growth, governments can and should be considering AI to help drive better outcomes for citizens.

The commercial advantages of AI are perhaps more readily apparent but it is the next generation of society level problems – the environment, anti-biotic resistant viruses, world hunger – that can be tackled through AI (alongside long-standing challenges such as innovation in transport, education and energy consumption). If the government wants to realise its ambition of putting the country “at the forefront of the artificial intelligence and data revolution” as outlined in the Industrial Strategy White Paper, then it needs to position the United Kingdom on a footing for success.

Firstly, the government needs to move quicker than other leading economies when it comes to the regulatory and policy environment surrounding AI. Public data sets – data relating to the general public but held by the government – hold massive potential for helping to meet some of the challenges outlined above. Imagine, for instance, visiting a GP and having any new medicinal prescription tested against your personal genetics to predict its efficacy, right there and then. Imagine a similar process for new cancer drugs.

Personalised medicine could become a reality, and unlocking the power of data could potentially be life-changing, but there needs to be structures in place to facilitate this in a secure, accountable and easily explained manner. The government needs to consider its policies regarding the use of public data to ensure that our society can benefit fully from the next wave of cognitive technologies.

Secondly, policymakers need to consider an often overlooked element of implementing AI at scale and that is infrastructure. Ensuring there is adequate technological infrastructure in place to interrogate the massive datasets and run complex AI algorithms is a prerequisite to being a world leader in AI. To carry out this work requires high-performance computer systems which are orders of magnitude more powerful than most organisations have at their disposal. The government needs a strategy to ensure that businesses and academics alike can access world-class computer systems to drive innovation and power AI so we can start delivering answers to big questions. Many countries around the world are already investing heavily in high-performance computing resources, so the UK needs a strategy in place to ensure it is not left behind.

Thirdly, a better link between industry, government and universities is required. The relationship between these three sectors needs to become more porous to ensure better cooperation, quicker failures (and consequential learning) and a clearer understanding of where resources are required and available. The government needs to consider a long-term strategy to tie these sectors together and build an ecosystem for closer collaboration – the Digital Catapult programme is a great start in this respect.

Finally, the dystopian man versus machine narrative is somewhat distracting in terms of what artificial intelligence will be able to achieve even in the coming decades. Instead we can think about augmenting human intelligence, for example the radiologist diagnosing cancer patients who no longer needs to examine every individual x-ray but can instead run an AI which will pull out only the at-risk cases for him to look at while he spends more time with patients. The data deluge and AI revolution are likely to create new jobs (a recent Gartner report suggested AI would create 500,000 more jobs than it eliminates by 2020) and possibly entirely new industries but a utopian prism is unhelpful when looking at the skills challenge. AI is not a zero-sum game but an understanding of where uplift is needed to close the skills gap to deliver on the government’s technological ambitions is required. Again, a closer relationship between industry, academia and government would be beneficial here.

We are generating so much information that it has already become impossible for us to sift through it. AI will help us understand the world around us and gain greater insight into our businesses, governments, countries and planet, but the work should be done now to ensure the UK can thrive in the age of artificial intelligence.

Matt Armstrong-Barnes is chief technology officer for AI at Hewlett Packard Enterprise.