Welcome to the AI revolution: From horsepower to manpower to machine-power

Before the invention of the internal combustion engine and the harnessing of electricity, humans weren’t the only members of the global workforce. Until the mid-20th century, horses were employed in the tens of millions across industries. In the USA alone their numbers reached 24 million, about as many as there are humans currently working in healthcare.

Also: How your business can best exploit AI: Tell your board these 4 things

“The draft animal population — the vast majority of which were horses and mules — grew six-fold between 1840 and 1900, from four to twenty-four million. This outpaced the growth in human population, which merely tripled during those same decades. By 1900, there was one horse or mule for every three humans in the United States. The majority of work animals lived and worked in cities and their surrounding hinterlands. The greatest uses of animal power were in agriculture and transportation,” from Animal Power by Anne Norton Greene.

Within cities, horses had provided most travel and transportation options for centuries. Apart from horseback riding itself, options included public transport by horse-drawn omnibuses, private coaches and carriages, and even ride-for-hire taxi services dating back as early as 1605 in London, provided by hackney carriages (four-wheeled) and later by hansom cabs (two-wheeled cabriolet carriages).

A shift from horsepower 

However, with the advent of the internal combustion engine, the number of literal workhorses in the US had fallen to six million by 1960. That figure has since fallen further to only about 1.5 million, of a total US horse population of about 10 million, most of whom are owned as pets or used in competition. 

The story is similar in Europe. In England, for example, just over three million horses were working at the beginning of the 20th century. That number had fallen below two million within a quarter of a century, despite the loss of human laborers to the First World War and the flu pandemic of 1918 to 1920 that took some 25 million to 50 million lives globally. 

A century later, in 2020, it is estimated that there are less than a tenth of that figure, around 160,000 horses, some 70% of which are pets and the rest are mostly engaged in racing and in some niche areas like mounted police and brewery dray horses. There are, in short, nearly no horses today in regular employment from a heyday of tens of millions fully employed.

Also: Time for businesses to move past generative AI hype and find real value

So, what happened? Automated movement happened. Until the internal combustion engine, we did not have reliable technology capable of carrying and pulling loads from one place to another, except by rail. We used horses and, for a while, the greatest existential threat to city living was the rapidly accumulating piles of horse dung. We switched to mobile technology, via the automobile, as soon as possible.

And now we’ve taken the next step. We are creating a whole suite of technologies that enable the automobile to live up to its name even more fully. “Auto” means self, and it originally implied a carriage (hence the word “car”) freed from the power of the horse. 

Now “auto” means being freed from the control of humans. It is technology by itself, autonomous, “under its own steam” so to speak. And in that sense, transport is starting to become something new. And the implications will be felt far beyond travel and transportation.

From manpower to machine-power

Until very recently, technology was first and foremost a tool. It was something humans built and then used to do a job — and to do it better, faster, and easier than we could without it. But still, we used technology. 

What’s new with artificial intelligence (AI) is that we are not creating new tools to help us do a job. We are creating a new workforce to do the job for us. This trend is not absolute of course and we can always point to older technologies that may have done part of our job for us (factory automation began at least 200 years ago). However, we are now creating a cheaper, faster, better, scalable workforce, not a cheaper, faster, better, scalable toolset.

This new workforce is not going to replace us all any time soon. There are two main reasons for this fact. The first is that the hype of AI far exceeds its current capabilities, except in some narrow, rules-based scenarios (e.g. games, in which it can far outperform even the greatest human players). 

Generative AI in particular appears almost magical in its ability to render text, images and even video. Yet its inability to understand any of its output, along with the volume of data and the power needed to train its models, surely limits it from replacing human workers. 

Also: When’s the right time to invest in AI? 4 ways to help you decide

That said, AI-powered capabilities are growing in orders of magnitude annually. By leveraging AI’s predictive and analytical capabilities, companies make informed decisions that benefit their bottom line, society, and the environment. 

However, new research shows only 30% of C-suite leaders feel confident in their change capabilities. Even fewer believe their teams are ready to embrace change. Lastly, 90% of IT leaders say it’s tough to integrate AI with other systems, citing two biggest challenges for AI adoption are data silos and application integrations. 

The second reason AIs will not replace humans any time soon is the time it takes our institutions to understand and embrace the capabilities of proven technology. We saw this most clearly in 2020 when school districts and businesses had to cease operations during the coronavirus pandemic because they had not yet implemented full online operations despite the capabilities being in existence for 15 years or more. 

We can expect late adopters to wait again until they’re presented with an existential threat before embracing AI and this lag will affect the whole. According to Accenture research, only 16% of the 1,000 organizations the consultant studied stand out as leaders capable of successfully managing the change necessary for adopting AI in business. 

Also: 4 ways to help your organization overcome AI inertia

Given those two major caveats, we can track the gradual integration of AI into the workforce and the eventual, inevitable reduction in the number of human employees as AI becomes cheaper, more efficient, and more accurate than us at performing a wide range of functions. 

It may be true that AI will create new opportunities that we can’t yet imagine, but they won’t be opportunities for more of us humans. In the short term, we may see more jobs, as not all technologies will develop at the same speed and will need our help to work effectively.

The machine-powered ecosystem

However, one day — and we expect that the real watershed moment will once again be fully autonomous mobility, from the vehicle to the android robot — the measure of manpower will become as figurative as horsepower is now. 

We will likely be startled by the manpower of the average robot. And we’ll start to see the emergence of a new measure of productivity — “machine power” or similar. This measure will be needed to represent how machines will no longer just do “human” jobs faster, more accurately, and cheaply. They’ll also be doing jobs that we can’t and are far more complex, with more inputs to handle, more moving parts to orchestrate, and less time to solve. 

Managing robotaxi fleets — the latest innovation in the centuries-old ride-for-hire service that no longer employs human drivers or horse “engines” — will be an early example of this new machine power. Managing fully autonomous companies will be another.

Also: Agile development can unlock the power of generative AI – here’s how

This transition will have large-scale societal implications, far beyond the scope of this piece. But to follow our logic to its conclusion, we humans will eventually go the way of the horse. There’ll likely be fewer of us, but we’ll live relatively healthier and happier lives. And once employment no longer sets our course, we’ll need to take seriously our one “job” of finding a new sense of purpose and joy.

Henry King, co-author of Boundless, and I are developing a framework for the different levels of capability that AI will need to demonstrate on its way to full autonomy in the workplace. In addition to the existing framework for autonomous driving, we were inspired by the SUDA operating model (Sense, Understand, Decide, Act) featured in our best-selling book “Boundless” and are incorporating this model into each level. We’ll be publishing that work on ZDNET soon.

This article was co-authored by Henry King, business innovation and transformation strategy leader and co-author of Boundless: A New Mindset for Unlimited Business Success.