by Frank Wammes

AI, robotics and the future of work

Nov 07, 2016
CareersEmerging TechnologyIT Leadership

AI and robotics could have a big impact on the job market. Many people have their opinion, but when and how will these technologies impact?

robot workers ts
Credit: Thinkstock

Robots will take over 6 percent of jobs in USA by 2021, says a recent report by Forrester. Oxford University is even going as far as to say that by 2035, up to 47 percent of our jobs are at high risk of being automated. If that number doesn’t have an impact, imagine this: by the time the current generation of infants legally reaches adulthood, they will have done so in a world where nearly half of our current jobs may be automated.

What are the driving forces behind this development? And which jobs will be targeted?

The driving force behind the automation of jobs – other than continually increased computational power, the heart of everything IT – are two separate developments: Robotic Process Automation (RPA) and Artificial Intelligence (AI). Often lumped together, it is important to distinguish between the two as they appear similar on the surface but are in fact vastly different.

RPA vs. AI

RPA is automation in purest form, working on an “if this, then that” basis. Mention the right word – or every synonym or misspelling you can think of – and the bot is triggered to fire off one or more of its many preprogrammed actions. It’s doing what we did before – only faster. This makes it easy to imagine customer service employees being automated – the customer calls with a certain complaint or question and the RPA gives the appropriate response. Process completed. It is RPA that will have the most notable, short-term impact on our jobs.

With AI, it’s a different story altogether. AI combines techniques such as machine learning and deep learning to constantly redefine its models and thus, to refine itself. AI is much more adaptive than RPA. RPA will do what you tell it to do, and it will do it again and again and again (when asked). AI is built to be self-learning and as such, it will change what it does as it learns and goes along. It is capable of learning from your behavior as it goes along. And this aspect, the ability to offer personalized, contextual information, will turn out to be more transformative than RPA could ever hope to be, even if it takes AI more time to come to fruition.

Our fear of a tyrannical AI is greatly surpassed by the ease with which AI slowly seeps into our lives.

And don’t kid yourself: AI is already all around us. The other day I got in my car and Google Maps told me: “it’s busy on the road, it’s a 57 minute drive to your home”. I was surprised, as I had not indicated that I was going home. But there’s no need to tell Google Maps that, as my behavior – driving home after work at a certain time of the day – is regular enough for it to guess where I’m going at that time of the day. Surprised at first, and perhaps a little agitated – “this is none of Google’s business!” – my reaction quickly shifted from agitation to expectation. As soon as I get in my car, I want to know how long it’ll take me to get to my house – and if Google Maps knows that traffic is particularly bad, why can’t it send a quick message to my wife (my next appointment) that I’ll be running late for dinner? Or, if we’re taking an Internet of Things perspective, it may even instruct my oven to cool down so dinner isn’t overcooked when I arrive ten minutes later than usual. Such is the power of AI, that it’ll be able to quickly shift feelings of surprise and intrusion into feelings of expectation and even feelings of need. There’s an undeniably addictive quality to it, which will ensure we bring more AI into our lives.

Already, every major tech player, and a whole legion of startups to boot, are developing their own versions of AI and the necessary data-gathering that comes with it. Through devices such as Amazon Echo and software such as Siri, Cortana and Google Now, massive chunks of data are channeled to cloud warehouses, waiting to be analyzed and put to good use. It’s inevitable that these developments will have a major impact on our economy.

Impact on our jobs

What then are the jobs that will be affected by either RPA or AI? Focusing on the different tasks that make up a job, rather than jobs as a whole, it is difficult to think of a job not impacted by either RPA or AI.

There are so many tasks, simple and complex, that are simply waiting to be automated. Of course, tasks that are repetitive are susceptible to RPA. In fact, as Oxford University puts it, all jobs that are ‘reliant upon storing or accessing information’, are also at risk of computerization.

Through big data and algorithms, AI is capable of finding patterns in the data that humans cannot find, even after years and years of education. Potentially impacted jobs include your mortgage advisor, your HR and marketing departments, but also your local physician.

Your AI physician may not know you personally as your human physician does, but consider this: your AI physician potentially has access to the world’s current knowledge of genetics and medication, is an expert in probability, is not subject to human biases and it knows your individual and family medical history – who’s going to make the better judgment? Again, these jobs may not disappear completely, but will certainly be affected in more ways than we can imagine.

On the other hand, what are the new jobs AI will create? The World Economic Forum created a list of characteristics required for jobs in 2020. These will be completely different than those currently required, with a greater emphasis on creativity and critical thinking. But these traits being the sole territory of humans is what will become obsolete. In a recent podcast, Marc Andreesen remarked that we always make the mistake of looking at the absolute numbers, not their relative impact. As an example, the estimate for 2016 is that about 24 million jobs will be eliminated in the US, which seems like a big deal. But what really matters is that 26 million jobs will be created as well.

The big question is: how are we going to re-engineer our society to cope with these massive changes to our current jobs? Certainly we will need to learn need skills, and unlearn certain others, but there are also economical, sociological and psychological consequences to consider. The progress of technology will also create new jobs, but inevitably, there will be gaps between what those newly without a job can (and want) to learn and what knowledge and skills the new jobs demand. But one thing seems certain: it is going to affect us all.