What keeps us busy in an age of automation?

Over the last 100 years, technological advancement meant a seasonal cry of the end of work as we know it. The robots will take all the jobs, the pundits would write and denounce with a misplaced Cassandra-like certainty. But the data tells a different story. If all the jobs are disappearing, then, why are we busier than ever before?

According to David Autor of MIT Stone Center on Inequality and Shaping the Future of Work the answer is simple: the perpetual invention of new work. Census data from the USA tell us that 60% of jobs sprouted into existence over the last 80 years, injecting fresh variety and depth into our daily grind.

“We create new variety and new depth to what we do.”

But what sets new work apart from old work?

Expertise.

Autor defined it as “domain-specific knowledge essential for achieving valuable objectives.”

As robots in whatever guise they appear in history start affecting job markets, there is growth in the number of lower and higher expertise jobs, while you see fewer jobs in the middle, which are usually blue-collar factory, administrative and clerical jobs.

“Why in the middle?”

“Many things that we do we didn’t formalise”.

Both so-called high and low skill tasks have elements of non-routine structures, which have stood in the way of being computerised, making them difficult to automate. On the higher end, “expert” work has been augmented through computerisation, by gaining access to higher levels of knowledge and information; while the “low skilled” work has generally not been augmented in the same fashion. Think of a policy writer having access to the world of information via the Internet, reducing the amount of time going to a library to perform research, versus a skilled janitorial worker who is, overall, using the same technology as fifty years ago.

However, the middle level of skill has undergone automation: call centres, self-checkouts, for example. This middle ground automation has generally pushed workers into the low skill side of the Barbell of Occupational Polarization.

Polanyi’s Paradox can help us explain this phenomenon.

“We cannot computerise tasks that we don’t explicitly understand.”Some forms of work can be readily codified and carried out by computers; things that follow well-understood rules and instructions which can be written as software, will be computerised. Middle-income jobs tend to require tasks that are easily automated. As these middle-income jobs become rarer, they push people into the low-skill end of the bar, where wages are low as you have an abundance supply of people who have knowledge or expertise to do this work. This happens even if the lower paid work is sometimes life-and-death situations and require a lot of expertise – say, a first responder role.

We are now in an era where computers can learn autonomously, accomplish tasks we would call creative, as well as do things we really cannot understand. (That Black Box system described before.)

So, what do we get out of all of this?

AI could commodify expertise.

AI could make a little expertise go further.

Autor sees an issue whereby an over-reliance on automated systems can degrade our own expertise, making us less capable.

The Handoff Problem

“If a machine does most of the work, it is very hard to pay attention”

To illustrate this point, think of the Simpson’s episode “King-Size Homer”, in which Homer replaces his menial task of pressing the Y key on his computer with a drinking-bird desk toy. If you know the episode, you know that nearly caused a nuclear meltdown.

If technology can be understood as something that shortens the distance between intention and action, what happens when an automated system gets in the way? Imagine the times when an automated door system doesn’t work, locking you out in the cold, and you have no idea how to fix it yourself. We want to use technology to augment our expertise, to compliment, and not replace – to make people more effective at what they do, not less.

Considering the current narratives about the trajectories of some automation technologies looming on the horizon, Autor’s thoughts help us navigate the storms brought in by the latest season of the end of jobs as we know them.

Download the full article and book: https://zenodo.org/records/14967269