A couple of years ago, I attended a talk by one of the world’s most renowned futurists. We all sat down to hear about how the world was going to be transformed by the merger of machine and human intelligence. And we kept sitting, because for 20 minutes technicians couldn’t get our expert’s PowerPoint presentation working. A vision of the machine-dominated future was held up because the technology of today wasn’t working properly.
It’s something I tend to think about whenever I read the latest study saying how artificial intelligence (AI) is going to destroy so many jobs and humans are set to become obsolete. Not because I doubt that technology will be transformative to the jobs and business models of today. But I do think there is a tendency to reach for extreme scenarios that often overstate either the extent or pace of change.
What I’ve learned in my time as CEO of insurer Simply Business is that technology in theory and technology in practice are two very different things. When I started out, my assumption was that we should try and automate everything, in the interests of efficiency. But when we tried to change the form people use to provide the information needed to generate an insurance quote – changing a lengthy document into one with just a handful of questions – the feedback was dreadful. People preferred the manual version in which they had to do all the work because they didn’t trust a quote that had been auto-generated based on proxy information.
That’s just one example. I also thought we would quickly automate our whole customer service function but it turned out people still want to talk to people and appreciate the human touch. So we have kept up a large contact centre team while using technology where we can to help people work more effectively and to automate some elements of the process like gathering customer feedback.
The point is that we tend to think in a very linear way about technology: the new way will be quicker and more reliable so of course it will be better. But as long as you are dealing with people as your customers you can’t make assumptions about how they are going to respond. Humans aren’t always as transactional as you expect them to be. Sometimes fast can be perceived as rushed, good value as cheap and efficient as impersonal. Technology isn’t some unstoppable force over which we have no control – it’s a tool that every business can and should decide how to use.
When it comes to the question of how AI and automation will affect companies and jobs, we need to be cautious about the forecasts that suggest seismic change in the near future. Without doubt there is going to be upheaval but it won’t be frictionless. Just look at how Tesla have struggled to make automation work for them in the manufacturing of its mass market Model 3 vehicle.
Whether it’s manufacturing, customer service or haulage, AI isn’t something that any business can simply plug and play. As I argue in my new book Reboot, it needs to be integrated in a way that delivers efficiencies within the context of a given company, which in many cases will mean finding ways of making machine intelligence dovetail with human.
The thing to remember about the apocalyptic, headline-grabbing predictions we often see is they are theoretical: models that have been developed based on perfect case scenarios of AI development and penetration. But no business is a perfect case scenario, and the evolution of AI within business is going to be an iterative process, not an overnight one.
So the next time you read a study saying that millions of jobs are at imminent risk of being automated, recall the last time you couldn’t get the Wi-Fi or office printer working. And remember that there is no such thing as an overnight technological revolution.