Artificial Intelligence – fact or fiction?

*This article first appeared in the Scottish Sun on 06/06/18*


“Skynet became self-aware at 2:14am, Eastern time, August 29th.”

A foreboding line from perhaps the most iconic portrayal of artificial intelligence in recent times – the Terminator film franchise, with Skynet as the true antagonist which seeks to wipe out human existence.

Ever since Mary Shelley wrote Frankenstein 200 years ago, fiction has been littered with man-made creations rising up and threatening humanity. Last year, Blade Runner 2049 depicted a bleak future with artificial intelligence – or AI – at its heart.

So it’s perhaps little wonder that debating AI causes people to have concerns.

But that vital debate is now happening, with the European Commission publishing a new Communication on Artificial Intelligence for Europe.

Questions are being asked such as: is AI a new phenomenon? Will AI systems replace human workers? Will robots take over the planet?

The dawn of AI was arguably in 1936, when Alan Turing proposed the concept of the ‘Turing machine’, a model of computation which triggered the development of informatics and computers.

It’s fitting that Holyrood will this week back Scotland’s own version of the ‘Turing Law’, pardoning all the men convicted of the crime of loving or just having sex with another man.

Turing himself, memorably portrayed by Benedict Cumberbatch in The Imitation Game, was convicted of gross indecency in 1952 and was pardoned posthumously in 2013.

After his codebreaking heroics during the Second World War, Turing went on to publish a paper on ‘computing machinery and intelligence’ that is often referred to as the origin of modern AI.

Over half-a-century on, and following huge technological advances, the European Commission (EC) today defines AI as ‘systems that can display intelligent behaviours by analysing their environment and taking actions – with some degree of autonomy – to achieve specific goals’.

AI is already part of our everyday lives. It can be found in Siri, or when you use Google Translate, or when you talk to an Alexa device.

But the economic impact of AI will be much more significant in the years to come and will bring about many opportunities for those who embrace it.

It’s estimated that worldwide business revenues could soar from 6.4billion Euros in 2016 to 37.8billion Euros by just 2020, making up a huge proportion of global Gross Domestic Product (GDP).

But this isn’t just about growing the economy. AI has the potential to achieve outstanding goals for humanity, from offering better and more personalised health diagnostics to increased road safety via driver assistance. The EC estimates that more than 90 per cent of all road accidents are caused by human error.

However, with any opportunity there are also challenges. And nowhere will that be more obvious that in the jobs market. Many new jobs will be created, yet many will disappear and many more will be forever changed.

Across Europe, countries must now start the work to enable people to acquire digital skills, through dedicated educational and training schemes. This should apply not only to students but to people of all ages.

There are diversity challenges to overcome: developers are mostly male and white, and people in more advanced countries will have an advantage from the outset. Policymakers have to set the right conditions to prevent AI technologies from leaving certain groups behind.

Automation is a word which politicians use more and more as they acknowledge the challenges, but few of those in power seem fully prepared for the seismic change that is coming.

According to the Commission, 37 per cent of the European labour force does not have even basic digital skills. Up-skilling and training will be essential in ensuring people embrace the transformative benefits which AI can bring. Both public and private organisations need to be on board for this to happen.

If we get it right, AI will enable people to be equipped for the world of the future.

At the European decision-making table, preparations are already underway. All those emails about GDPR that filled your inboxes were part of this readiness programme.

Were the emails annoying? Yes. Were they a vital step? Yes. AI systems need very large quantities of data to be either provided to them or collected by them.

Access and storage of data could be a major issue across Europe, but the GDPR introduces several guarantees, protecting individuals and strengthening the roles of consent for the processing of personal data.

It’s this kind of innovative thinking that Britain will miss if we go ahead with the reckless plan to leave the EU next year.

With the right ethical and legal framework, Europe can become a global leader and trendsetter when it comes to AI.

But will robots eventually take over planet Earth? The answer is no. Away from the science fiction world created in films like The Terminator, data-driven AI systems can only perform one task at a time and cannot transfer their knowledge to another task, as humans can.

Once they are programmed to do one task, that’s essentially what they do, nothing more.

Alexa is not going to become self-aware. Well, not yet, anyway…