Will AI do to professions what the Model T did to railroads?

by Rita Gunther Mcgrath: Associate professor at Columbia Business School.
As economist Carlota Perez so eloquently lays out, systemic changes in technologies always lead to systemic changes in society. The winners in an old regime become the losers in a new one. With many predicting that the AI revolution will democratise access to expertise, what changes might we expect?

Technological advances make the formerly impossible possible, sparks excitement when their potential is realised and then come to be taken as utterly ordinary. By the time technologies have become so diffused that they are part of daily life, they will completely rewire the societies in which they are commercialised.

Railroads, for instance, with their ability to transport goods at previously unprecedentedly low rates made the canal-and-turnpike system that preceded them obsolete and gave rise to robber barons and steel magnates.

This is the thesis of many economists who study long cycles of development in capitalism, famously Joseph Schumpeter and more recently Carlota Perez. Schumpeter put the technological entrepreneur at the centre of his theories of economic change, noting that each successive “wave of creative destruction,” as he put it, rendered obsolete many of the solutions from a previous era.

Carlota Perez builds on Schumpeter’s idea by arguing that the booms and busts of capitalism are a feature, not a bug. Lured into exciting new technologies by their promise of great riches, capitalists flood the new sectors with money and encourage entrepreneurs to break down the old systems and create something new. Eventually, as the new technologies are more widely adopted, societies change to reflect the new economics of what is now possible. Winners in the old regime become losers in the new one and resist it. Eventually capital comes back into the “production economy” and what was once considered an impossible fantasy becomes ordinary reality.

The Model T and mass production

Consider the world in 1900. One got around on foot, by horse or by using a railroad or steam ship. The first immense modern corporations grew up, fueled by railroad and steel monopolies. The Rockefellers, Vanderbilts and Carnegies took over from the people at the top of the economic ladder. Although the first automobile, the Benz-Patentmotorwagen was invented in 1886, it cost $1,000 (about $33,500 in today’s dollars), and only about 25 were built.

By 1901, William Maybach of Germany designed what is widely recognised as the first modern motorcar, the Mercedes, an expensive product that performed at a high level. In parallel, Ransom Olds in the US created what was essentially a glorified horse and buggy that sold, however, for a lower price point, though neither offer was a true mass-market product.

It took Henry Ford’s adoption of the assembly line (inspired by the practices of the butchering business) to create the conditions in which cars could be produced at an unprecedented scale for a remarkably affordable price. Among Ford’s other innovations were the introduction of a living wage ($5!) for factory workers (reportedly so that he could keep them working in the boring factory) and a system of auto financing which let customers pay for their vehicles over time. Cars started to be sold in mass-market numbers, putting pressure on systems that were designed for the previous technology.

Freed from the constraints of living near railroads or other transport hubs, the automobile allowed the development of spread-out cities. In the years between 1910 and 1940 most Americans could aspire to owning a car. Farmers started to use cars to transport goods. Even as cars (and trucks) become more popular, reliance on railroads as the primary way to transport goods dropped. A “good roads” movement emerged in that period to replace dirt roads with their rough surfaces, tendency to get muddy and dusty with superior paved roads, eventually being superseded by the creation of the Federal Highway system in America in 1926. Cars eventually were sold by the millions and new players - Fords, Mellons and Waltons - rose to the top of the economic ladder.

Social arrangements were utterly changed as spread-out suburbs post World War II created what many look back on with a sense of nostalgia. Capitalism seemed to have delivered on its promise to use productivity gains to improve the quality of life for a great many people, giving them the potential to hope for a better future. Perez notes that this is characteristic of a “golden age” when the technology is becoming widely adopted and its productivity benefits are now seen.

Unintentional de-skilling

Further innovations meant that manufacturing processes became standardised, allowing the production of all sorts of goods without skilled craftspeople. Innovations piled upon one another, creating new jobs and possibilities but also destroying old ones. Electricity wiped out lamplighters, for instance. But society changed in other ways - mass-production meant that a person with a general-purpose education could be productive.

One of the unintended consequences of these kinds of changes is to “de-skill” certain kinds of tasks. Where possible, new techniques meant that the output of craftsmen, machinists or factory workers could be performed more cheaply by people with less training and practice. What this means in practice is that equivalent tasks can be performed by less expensive people.

Such deskilling has mostly been thought of as taking tasks away from physical jobs. What we’re starting to hear lots of chatter about, however, is that AI has the potential to lead to the deskilling or democratisation of knowledge-based jobs.

The democratisation of expertise and what that means for knowledge professions

David Autor, the MIT economist, has hinted at some of the implications of AI for jobs requiring formerly scarce expert judgment. As he puts it, “The unique opportunity that AI offers to the labor market is to extend the relevance, reach, and value of human expertise. Because of AI’s capacity to weave information and rules with acquired experience to support decision-making, it can be applied to enable a larger set of workers possessing complementary knowledge to perform some of the higher-stakes decision-making tasks that are currently arrogated to elite experts, e.g., medical care to doctors, document production to lawyers, software coding to computer engineers, and undergraduate education to professors.”

As more parts of the global economy have rewarded knowledge work, expertise has come to be expensive and highly rewarded. As Autor puts it in a different article, “Consider the occupations of air traffic controller and crossing guard. In broad strokes, these are the same job: making rapid-fire, life-or-death decisions to avert collisions between passengers in vehicles and bystanders. But air traffic controllers were paid a median annual salary of $132,250 in 2022, or nearly four times the $33,380 median annual pay of crossing guards.” The reason is that it takes long years of practice to become an air traffic controller - the expertise is scarce, therefore it requires a price premium to repay that investment.

What AI does that previous generations of computers could not do is to respond in novel, not pre-programmed ways to stimuli. This opens up the opportunity that people with less training and lower skills are able to produce the outputs (in many cases, choices or decisions) that would have previously required extensive training and field practice. Autor suggests the consequences, “By providing decision support in the form of real-time guidance and guardrails, AI could enable a larger set of workers possessing complementary knowledge to perform some of the higher-stakes decision-making tasks currently arrogated to elite experts like doctors, lawyers, coders and educators.”

So what may all this mean for the consulting profession and others? Well, with answers to key factual questions available at the push of a button, that form of expertise is likely to become far less pricey. Unique research results (based on surveys, for instance) may be less valuable than conclusions based on actual behavioral data. As AI incorporates models that let people practice and get feedback, even the “reassurance” function of a consultant may become less valuable. And it is only going to get harder to protect an advantage based on any kind of data in a world where it is swirling all over the place.

Just as the convenience and lower cost of transportation sparked by the Model T decimated the railroad business, the convenience and lower cost of decision support made possible by AI could cause the bread and butter of the consulting business to erode.

Useful resources:
Rita Gunther McGrath
Rita McGrath works extensively with leadership teams in Global 1000 companies who wish to develop their capability to drive growth.
Share on Twitter Share on LinkedIn Share on Facebook
Share via Email
©2024 SURREAL. All rights reserved.
Follow us on Twitter Follow us on LinkedIn Join us on Facebook