In 1915, there was about 21.5 million horses pulling carts, moving people, rustling cattle and doing all kinds of other jobs in America. A hundred years later, there are less than 2 million and very few of them do jobs. An American software engineer, Gerald Huff conducted a comparative analysis in 2014 of occupations in the U.S. between 1914 and 2014. The insight was that 80% of the occupations of 2014 existed in 1914. So while the American economy is much bigger, the net gain of new jobs is quite small. There are those that fear digital technologies like Artificial Intelligence and Blockchain will replace humans entirely or mostly. Perhaps. Or perhaps not. We just don’t know and making any definitive statement would be foolhardy at best.
Many of the jobs in the near future we can’t name because we don’t have words for them, since the technologies haven’t yet been invented.
In the meantime, we can look at how humans will work with digital technologies over the next couple of decades. Essentially, how we will or can, classify jobs. I’ve come up with four layers of jobs, starting with full automation where humans are not at all involved, or simply “in-the-loop” for key aspects, such as pulling the plug if something goes sideways. Like warfare, stock trading, medical procedures or energy systems infrastructure.
The diagram below is a visual representation of how these jobs may look in the near future, which is the next one to two decades. Taking into consideration that we are in the Exponential Age and technologies are evolving at an exponential rate that it is hard for us to predict just how they’ll evolve.

The Four Layers of Digital Work Macro Classifications
- Anthro: While they will rely on layers of support from both technology and humans, the human makes the critical decision. This could be a Prime Minister or government minister, a general or field commander or a judge or a CEO or other senior executive who will have to weigh the inputs from AI tools or analytics but ultimately needs to make a human-centric decision.
- Augmented: In this classification, the human does 60% to 80% of the work. This could, in some cases be a doctor or surgeon who gets support from an Artificial Intelligence application or again, a lawyer or judge who gets analytics or AI support, but must make the decision. This may also be a creative director in a marketing agency that has insights, but makes the final call on the creative execution. Or a product manager or UX strategist.
- Centaurs: Half human, half machine. Essentially, machines can do 50% of the work alongside a human. In knowledge industries, this could be an insurance clerk who makes a final decision on a claim, or a lawyer who writes up the final argument for a litigation case.
- Automatons: Largely, up to 99% automated with minimal human-in-the-loop actions that are restricted to defined emergencies only, such as a surgery and the doctor needs to step in when something goes wrong.
This is of course, conjecture. A framework for considering how we will have to adapt alongside the technologies we are developing. We are entering the era of Cognitive Adaptation and the Exponential Age. Where technologies accelerate at exponential rates unlike we’ve ever seen before and whereas before, humans largely used physical tools to adapt, we are now having to cognitively adapt. There are risks and dangers, for all technologies are double-edged swords. A bronze axe many millennia ago could feed a family and build a shelter, it could also be used as a weapon to smite ones enemies. So today, can technologies like Artificial Intelligence, robotics, drones and even gene splicing be used for both good and bad.
Setting out a framework such as this however, can help us to navigate how we will adapt to work and economies. To decide where and how information technologies should be deployed within organisations. It may also help through transition periods where massive amounts of humans will be replaced by technology. Such as deciding when people should get a Universal Basic Income, re-skilling and re-deployments.
Leave a Reply