In recent decades, technology has transformed the nature of many occupations, leading to a rise in highly paid jobs while diminishing the number of middle-skilled positions. The impact of artificial intelligence (AI) on the future of work, however, may differ and could potentially revive the struggling middle class.
"God created men, and Sam Altman made them equal" - this famous phrase about inventor Samuel Colt, who introduced affordable revolvers to the market and forever changed the world, could be paraphrased to describe the potential of AI. Many people associate AI with ChatGPT from OpenAI, a company led by Sam Altman.
A study by economist Erik Brynjolfsson and MIT economists Danielle Li and Lindsey Raymond suggests that AI could reduce the gap between novices and experienced workers in certain routine occupations. The research focused on the use of AI tools in customer service departments of an undisclosed large software company that provides administrative programs to smaller firms. The study's findings indicate that AI is most beneficial for beginners and less experienced workers.
The monitored employees utilized an early version of the ChatGPT system for their work. This means that the employees were using a significantly less capable system than the language model that has captured the public's attention in recent months.
Nevertheless, the researchers concluded that employees who used AI were, on average, 14% more productive, able to handle more customer requests within an hour. These workers also reported higher satisfaction, and the turnover rate, especially among newcomers, decreased. Customers who interacted with AI-powered employees also experienced increased satisfaction.
However, AI did not equally benefit all workers. The impact depended on the employee's level of experience. Highly skilled and experienced workers prior to AI implementation benefited very little or not at all. On the other hand, significant improvements were observed among novices. "The system demonstrated that people with just two months of experience achieved performance levels comparable to those with six months of practice," described one of the study's authors, Erik Brynjolfsson, in an interview with NPR.A closer look reveals the logic behind these findings - AI learned from the data and behaviors of the most skilled and experienced employees and then imitated their actions.
This differs significantly from the impact of previous technological waves on the job market and the nature of employment. According to researcher David Autor and his colleague David Dorn, the proliferation of computers and the internet in society had the effect of benefiting highly educated workers while reducing less skilled positions, which could be filled by individuals with less education.
"Computerization does not reduce the quantity of jobs but rather worsens the quality of jobs for a significant portion of workers," wrote the scientists in a commentary for The New York Times. Consequently, workers without a college education gravitate towards manual labor jobs, such as food services, cleaning, and security, which are numerous but offer low wages, uncertain job security, and limited prospects for advancement, according to the authors.
One of the study's authors, David Autor, now believes that AI could reverse this trend and save the middle class. However, he emphasizes the need to view AI as an assistant to workers rather than a replacement. "Although it is unlikely that artificial intelligence will take away your job, it will certainly change it," summarized political theorist Tom Parr.
Another aspect of AI's emergence is the removal of barriers that certain professions possess. A typical example is the IT sector, which requires a considerable amount of time for individuals to reach a high level of expertise, with numerous specialized fields within it.
Some refer to this change as the "Uber moment." When the ride-sharing service Uber emerged over a decade ago, knowledge of precise street locations in a given city became unnecessary overnight, as anyone could easily determine them using smartphones. Consequently, taxi drivers in London experienced a sudden drop in revenue, up to 10%, as Uber opened up the livelihood to a wide range of individuals. This led to the "juniorization" of the industry, replacing more expensive and experienced workers with younger ones who could compete with them using technology and offer their services at lower prices.
However, it is clear that AI cannot completely replace humans. Automation comes at a cost, and as Europe, in particular, is expected to struggle with energy shortages in the coming years, it will become increasingly challenging. "Artificial intelligence consumes more energy than other forms of computing, and training one model can consume more electricity than 100 American households in a year," according to Bloomberg.
In some fields, it will still be cheaper and possibly more sustainable to rely solely on "dull" human labor. "A machine could do it, but it would be quite expensive," responded an individual responsible for warehouse automation at Zásilkovna when asked about the possibility of using robots for all tasks in the warehouse. If we keep our fingers crossed, AI may not spell the end of the world for us office workers. However, we will have to wait and see how it unfolds, as artificial intelligence is still in its early stages of development.