Generative AI tools promise to increase the productivity of white-collar workers. Along with more traditional AI and machine learning tools, today’s GenAI is already helping us with diagnoses, contracts, drafting, translations, transcriptions, and more.
But what might its impact be on income inequality? It’s not what most people expect, in my view. And that’s a good thing.
Consider: Rising income inequality in advanced economies has been one of the most disruptive elements of globalization and automation over the past three decades. Ever since information and communication technology (ICT) gave new, significant advantages to already-advantaged workers with university degrees, increasing income inequality has been undermining social cohesion – and even unleashing violent protests and paving the way for populist leaders.
However, the advent of GenAI presents a different scenario. With this technological advancement, I foresee a new skills twist favoring less-skilled workers, giving them advantages that will help level the playing field. People who think technological progress will only widen the gap between rich and poor are failing to consider what happened before 1970, as I’ll briefly explain.
“From 1970 onward, tech advances – mainly via computers – began to drive a wedge between white and blue-collar workers. The invention of the computer on a chip in 1973 was the cause, as university degrees became prerequisites for utilizing ICT effectively.”
From 1870 to 1970, technological progress in the form of mechanization – with engines, electric drills, tractors, and so on – was essentially equalizing for workers in advanced countries. Mechanization gave more power to the people who worked with their hands, whether on farms or in factories. At the same time, it had much less impact on those who worked with their heads. As machine-driven productivity gains led to higher wages for lower-income individuals, inequality was compressed. This period, sometimes called the Great Compression, saw income inequality fall across nearly all countries in which we have data to analyze.
However, from 1970 onward, tech advances – mainly via computers – began to drive a wedge between white and blue-collar workers. The invention of the computer on a chip in 1973 was the cause, as university degrees became prerequisites for utilizing ICT effectively. So, those who were already more skilled and earning higher wages gained the most from ICT, widening the wage gap over unskilled workers. At the same time, unskilled workers found themselves competing with and replaced by robots and automated processes. The rising productivity of white-collar workers combined with computer-enabled substitutions for blue-collar workers is what’s known as the (first) skills twist.
“Robots have already replaced humans in steel-collar jobs, such as street sweeping and assembly-line work.”
But AI is twisting skills in a new way: It commoditizes experience, training, and education – hallmarks of society’s most highly skilled workers. To put it starkly and simply, machine learning is data-based pattern recognition. What high earners – people at the top of their professions – generally rely on is experience-based pattern recognition. Now, whenever experience can be reflected in data, it can also be captured by AI-trained models and reproduced costlessly, perfectly, and endlessly. With AI, accumulated professional experience diminishes in value. The playing field is being leveled.
Meanwhile, this new skills twist does not threaten manual laborers, as most mechanical automation has already occurred in G7 and similarly advanced countries. Robots have already replaced humans in steel-collar jobs, such as street sweeping and assembly-line work.
Instead, today’s GenAI is targeting white-collar professions, but it augments instead of replacing skilled workers. For example, copy editors checking grammar, spelling, and adherence to a specific professional style can now rely on GenAI as a shortcut to years of training and experience. Human eyes are still required to correct AI shortcomings, but one person using AI tools might take on the work of an entire team of editors from the pre-AI era, freeing up time to tend to other skilled tasks.
“Consider a nurse in a hospital working with an advanced medical AI to diagnose patient issues, develop treatment plans, and identify potential conflicts in treatment.”
I see the latest skills twist introducing a new social dynamic where automation actually benefits less-skilled workers more than their highly skilled counterparts.
Consider a nurse in a hospital working with an advanced medical AI to diagnose patient issues, develop treatment plans, and identify potential conflicts in treatment. This nurse should become significantly more productive, taking over tasks that doctors used to do. A doctor using the same AI experiences a productivity boost, too, but it’s not as pronounced. As a result, AI will likely create middle jobs between highly skilled professionals like doctors and their less-skilled assistants who are able to use AI effectively.
There’s room for more middle jobs to be created between the ranks of lawyers and paralegals, engineers and road chiefs, architects and drafting technicians. In all these professions, AI can enhance the productivity – and income levels – of a new class of middle-skilled workers.
In this scenario, the middle class can flourish again, while the highest-skilled, highest-income individuals will experience modest gains. Meanwhile, a very small group of individuals controlling these technologies may amass mind-boggling wealth, but their billions (or even quadrillions) will be less relevant to the broader workforce’s well-being. Think of them like the world’s top soccer stars. As long as most people are doing better, who really cares about, say, a dozen people’s wealth?
Along with new middle jobs, there will also be a reshuffling of employment. Tech advancements always come with some reshuffling. That’s the way the market works. We saw it as we moved from farm to factory, from factory to office, and from office to ICT-enhanced office. That said, there will still be jobs for all. Post-AI, jobs where experience is harder to turn into training data include motivating people and managing large groups. We still haven’t seen those skills in an app, so there may be more work as coaches, educators, and other areas as determined by market forces.
“GenAI can reverse this trend by augmenting the productivity of less-skilled workers in advanced economies.”
So far, the business model for AI development mirrors that of Microsoft with its Office products: relatively low prices coupled with wide distribution. This approach ensures broad access to productivity-enhancing tools, spreading the benefits across many users. Although the leading AI developers may hold near-monopolistic positions, I see their pricing strategies remaining accessible, in keeping with Microsoft’s model.
As today’s AI primarily targets service workers in New York, London, and other places with lucrative markets to tap, its initial effects will be felt in these advanced economies, with secondary impacts eventually reaching developing nations. However, while developing economies have larger agricultural and informal service sectors, AI’s transformative potential is less immediate.
At the end of the day, while the last few decades of technological progress have increased income inequality and caused extreme disruption, those who predict more of the same are simply wrong, in my view. GenAI can reverse this trend by augmenting the productivity of less-skilled workers in advanced economies. This shift will foster a more equitable distribution of income and promote more social cohesion – reasons to be optimistic about an AI-enhanced tomorrow.
AI x 9: This article appears in a 9-part summer series that examines how AI is impacting leadership and business, produced in collaboration with Expansión.
Leading consumer goods companies are strategically embracing AI across their entire value chains, creating unprecedented business capabilities and market differentiation. IMD’s AI Maturity Index reveals the strategies that separate AI innovators from...
AI and sustainability are transforming the business landscape. In their new book, Michael Wade and Konstantinos Trantopoulos introduce the concept of twin transformation – a leadership imperative for companies seeking to remain...
AI's true revolution isn't in chatbots but in transforming industrial operations through dark factories and predictive maintenance. This will require leaders to balance technological advancement with responsible societal impact....
In more mature stages of Artificial Intelligence (AI) adoption, AI agents are an integral part of business operations, bringing new layers of complexity and enormous opportunity, says José Parra Moyano....
Explore first person business intelligence from top minds curated for a global executive audience