I recently enjoyed the series of TV programmes on Sky News and the BBC profiling developments in Artificial Intelligence (AI) and robotics. But while the subject has always fascinated me, what really struck me was the underlying theme of fear that pervaded both programmes: the fear of what advances in AI might mean for the job security and the future of humans. There seems to be a war-like philosophy – if we develop them, they will overtake us, putting our very existence in danger. Robots will take over the world, and we will lose control.
Discussions and research about the use of robots can be dated back to the first millennium, but the conversation today is indeed a world away from the musings of days gone by. Whereas we used to think robots' potential was limited to repeatedly performing the same process-based tasks (typically on factory floors), we are now creating machines that can replicate a human’s ability to think, making decisions independently based on patterns, past performance and logic.
This phenomenon is a subset of AI known as machine learning. Whereas the robots we think of on those factory floors are programmed to perform, machine learning was defined by pioneer Arthur Samuel as "a field of study that gives computers the ability to learn without being explicitly programmed".
The robot in our pocket
Although the process has progressed slower than many anticipated, the concept of machine learning is now steadily infiltrating our lives, and millions benefit from it daily. Google Search, used daily by around 300 million people, uses machine learning. Every time a term is searched and a link clicked on, Google's search engine learns from the user’s behaviour. It then hones what is displayed in the search results on subsequent occasions. To do this, complex algorithms analyse multiple behaviours in a bid to create a more personal response for the individual.
On our mobile phones, too, we see AI in action. Intelligent personal assistants such as Siri and Cortana can take a stream of speech, interpret it and offer solutions by individualising responses and delegating to other applications to complete tasks. Our media consumption has similarly been transformed. Netflix, Spotify, Amazon and Apple all use machine learning to make recommendations to customers of films, programmes, albums and books they might like based on their own choices and those of other customers.
While these applications of AI have been widely accepted, other developments are causing more of a mixed reaction. Driverless cars are hitting the headlines and grabbing the attention of everyone. While advocates try to reassure people that removing human error can make driving safer, sceptics are terrified by the prospect of cars "driving themselves" around our cities and countryside. Can a car be programmed to act more safely than a driver would, to take into account the magnitude of variables including other humans' behaviour, weather and the environment? Science says yes, but this is a concept that many humans find difficult to accept.
AI penetrating banking and telecoms
In my own industry, financial services, machine learning has a multitude of applications, some of which are already well-established. One example is the use of complex machine-based algorithms during credit card transactions to predict the chances of a transaction being fraudulent. To do this, banks use a finely developed understanding of user behaviour patterns – saving millions in the process.
The appetite for using AI in financial services and telecoms has shown exponential growth in recent years, and shows no signs of slowing down as banks and mobile phone operators seek new ways to cut costs and maximise customer convenience.
We can laugh or wince, but all of these developments will iterate further. A rise of the machines is inevitable. Afraid?
Fear not: imagine the freedom
For many people, the prospect does not sit comfortably. Concerns range from the immediate to the apocalyptic. If machines take over all our jobs, how will we earn money? What if machines develop so much that they overthrow the human race, Terminator-style?
I cannot help but think this is an extremely short-sighted view. The reality of AI integrating our everyday lives in this way is hundreds of years away, and the landscape in this imagined future will be very different from the world we enjoy now.
So what if we turn the question on its head? If robots are going to be doing all the work for us, performing all the tasks that look after the economy and drive through innovation, what else could us humans be doing? Why do we fear a time where we are not tied to a computer screen or breaking our backs with manual labour, and instead have time to experience the wonders of the world we live in? Can anyone who owns a dishwasher really say that they miss the time they used to spend washing dishes? Do car owners long for a horse and cart to make the 30 mile commute to work every day?
Or maybe we could spend our time putting right the wrongs that our generation and the ones before it have made. It seems bizarre that in a time when our news headlines are full of the extremities of war and the impact this has on human life, as we sit in our offices reading about families fleeing for their lives and putting themselves in extreme danger because of the actions of other humans, that we fear our species’ risk is from mechanical devices rather than people. As a race we have the opportunity to nurture the growth of AI, such as a parent does with a child, to create a future where both can co-exist. If we get this right, our robots will enrich, rather than endanger, our lives.