AI is coming


“It is difficult to make predictions, especially about the future”. [Niels Bohr]

We are all aware that progress is being made in the field of artificial intelligence (AI). There are already AI applications across a variety of activities, including entertainment, education, health, commerce, transport, and services, to name some.   

If you were to ask your colleagues whether they think AI will prove to be a net job killer or a net job creator, chances are you’ll get a variety of opinions.  

On one hand there are those of us who feel that AI is likely to trigger a process leading to widespread loss of employment; mild, severe, even catastrophic. For others, AI is a potential enabler, helping to increase productivity, making life easier.  Still others will say that some jobs will lose, while others will gain… and so on.

Such opinions are mirrored in the wide range of predictions analysts are offering: Massive net job losses vs net job gains, and a range of scenarios in between. No one seems to be sure.

The basic definition of AI seems simple enough. Here’s one: “AI is the ability of a computer program or a machine to think and learn.” (Thanks, Wiki!)

Yet, not only is there debate on what “think” or “learn” actually means, (as opposed to, say, “mere “automation), there is added fogginess when it comes to perceptions as to AI’s role as an eliminator or an enabler. As a thought exercise, consider: Is Siri a net enabler, or threat to jobs? What about driverless trucks? Enabler or threat? If AI was incorporated into street lights – enabling, or disabling? Or an automated robot already on a conveyer belt, now programmed to learn as it goes, how does that affect jobs?

What is certain is that there are divergent approaches used to prepare and analyze the information. And that results in widely differing predictions.

On the downside, are reports predicting losses – and they are most widely referenced:

A 2013 Oxford study estimating that 47 percent of all US jobs could be automated by about 2030.

A study commissioned by the Organisation for Economic Co-operation and Development (OECD) in 2016 concluding that 9 percent of jobs in the organization’s 21 member countries would be automated within a decade.

A November 2017 McKinsey report on “…workforce transitions in a time of automation” suggesting that 400 million to 800 million jobs worldwide could be automated by 2030.

In opposition are studies which argue that while AI will eliminate the need for many different kinds of jobs, in many different categories, it will at the same time also create new jobs in categories that we know of, and many more in categories that have yet to be created.

One such view, interesting in its approach, is that of Brynjolfsson and Rock, (MIT), and Tom Mitchell (Carnegie Mellon University). They argue that we’re having the wrong debate when it comes to AI: instead of pondering whether jobs will be wiped out or not, and which, and how, we need to focus on the redesign of jobs and re-engineering of business processes to incorporate AI. For them it is inevitable: “While AI and machine learning will be everywhere, the suitability for machine learning of work tasks varies greatly. The high and low suitability-for-machine-learning tasks within a job can be separated and re-bundled.”  (ref Forbes: Artificial Intelligence Will Replace Tasks, Not Jobs)

So instead of worrying about job losses, executives should be helping to reduce jobs in which AI and machine learning take over boring tasks, while humans spend more time with higher-level tasks.

To that I say “good luck”. I’m going to stick my neck out on this one. Yes, I think that there are areas where AI will be a positive, for a range of activities, across a wide range of enterprises, in a variety of sectors. However, there is a real risk that AI will threaten employment on a wide scale. We have time now to find ways to avoid that, and we should.

If the advance of AI is inevitable, then let those who can, start looking critically at how and where AI should be used, and where it needs to be limited and regulated. Anathema to some, I know, but unless we are mindful about where we allow AI to go, we could be heading for job insecurity on an unprecedented scale, and all the misery and mayhem that might come with it.

Author: Kevin Abraham