A 19th-century economist has been capturing attention lately. William Stanley Jevons, the British marginal revolutionary, concluded 160 years ago that when technology improves the efficiency of a resource, demand for that resource can increase. Today, many people are worried that artificial intelligence (AI) will wipe out jobs, but employment levels in AI-sensitive industries are reaffirming the Jevons paradox, also known to some as the Jevons employment effect.
Jevons Paradox in the Real World
About every generation or so, futurists warn that some new technology will transform industries, replace human workers, and lead to a dystopian future. Automobiles, computers, and AI – the doom-and-gloom viewpoint becomes conventional thinking and popular narrative. Eventually, time passes, everyone becomes wealthier, and we forget about these predictions, moving on to the next panic attack. The reality, however, could be what Jevons discussed: efficiency reduces costs, fuels greater demand, and leads to higher consumption.
For years, radiologists were supposed to be wiped out. In 2016, early deep-learning models prompted the so-called Godfather of AI, Geoffrey Hinton, to urge the world to stop training radiologists now. Since then, not only has the number of radiologists jumped – from about 31,000 to almost 38,000 – but average income has also nearly doubled to around $550,000.
While online booking has been around for a while, AI was supposed to be the death knell for travel agents. Employment growth flatlined from 2010 to 2019, but payrolls have surged since 2021. Cheaper travel booking led to greater total travel demand, with firms hiring human agents for intricate, exotic adventures in Tierra del Fuego or corporate accounts.
In the Philippines, roughly 2 million people are employed in call centers, rising every year since 2016, despite the pandemic, trade strife, and the AI revolution. Chatbots are ubiquitous, but companies are still investing in the human experience. Torsten Slok, chief economist at Apollo Wealth Management, put together the best explanation: "Lower cost per interaction does not mean fewer interactions. It means more customers served, more channels opened and more markets worth reaching. The technology that was supposed to shrink the industry is fueling its expansion."
The software scare, also known as the SaaSpocalyse, has hammered tech stocks. The latest worry is that AI, such as Anthropic’s exceptional Claude, is disrupting conventional business models. As a result, software development jobs are heading the way of the ice man. However, job losses in software have been much smaller than feared, and Indeed job postings have been on an upward trajectory since October.
Tech Complements, Not Erases
The next decade will reveal whether AI truly displaces work, and billionaire Elon Musk’s vision of a universal high income comes to pass. The questions now looming over the economy – which skills will matter, which jobs will surge in demand, and how the broader system will adjust – will all have answers heading into the 2030s.
Will some positions be swept up and dumped into the dustbins of history? Absolutely. Adaptation will be paramount. In this case, the next question becomes: Will new AI-fueled jobs be created faster than the ones it kills? This is where the government school system will likely struggle to adapt, leaving the private sector to bear the brunt of education and training the incoming workforce.
For the current crop of jobs, attorneys can put together contracts faster, doctors can diagnose more accurately, accountants can navigate the complex tax laws more easily, and reporters can research more efficiently. If history is any indication, AI, robotics, and machines will complement blue- and white-collar work instead of forcing us to lead an idle existence on the island of Ogygia.
Jevons paradox, or how the world learned to stop worrying and love artificial intelligence.
