Advertising
For help, call:
About a decade ago, Carl Benedikt Frey and Michael Osborne, two economists, published an article that went viral. They claimed that 47% of U. S. jobs were in danger of being automated. An avalanche of studies followed suggesting that the poorest and least informed were the most vulnerable to the coming revolution. These fears have intensified as synthetic intelligence (AI) functions have advanced. On Nov. 2, speaking after the U. K. AI Summit, Elon Musk predicted, “There will come a time when there will be no more jobs. “necessary. “
But, at the same time, economists are more optimistic. Recent studies have shown that fewer staff are exposed to automation than MM. Frey and Osborne (see Figure 1). In 2019, Michael Webb, then of Stanford University, showed that AI patents are intended more for professional jobs than for software and robots. The new AI seems better when it comes to coding and creativity than anything else in the physical world, suggesting that unprofessional jobs may simply be siloed. In March, Shakked Noy and Whitney Zhang, both of the Massachusetts Institute of Technology (MIT), published an experiment that appeared to increase productivity more when writing low-capacity staff than higher-capacity staff.
Although AI is still in its infancy, some industries have enthusiastically adopted it. A close look at three of them (translation, visitor service and sales) largely supports the economists’ optimism, not without complications. In the translation industry, perhaps the first sector greatly affected by language modeling, staff have become reviewers, putting order into a first draft made through AI, making it less difficult for beginners to enter to the industry. In visitor service, AI has helped the functionality of newcomers. In sales, the most sensible players are using generation to locate potential customers and take notes, distancing themselves from their peers. Will AI increase the incomes of superstars more than those of laggards, similar to the web revolution? equalizer”, increasing the incomes of the poorest but not those of the rich? The answer would possibly depend on the type of task in question.
Roll the dice
Roland Hall has been translating board games and commercial fabrics from French to English for 27 years. Remember that back in the 90s, software was used to translate express words from one language to another. Nowadays, equipment is more advanced, which means that the types of jobs that can be done have been divided into two. One type includes texts where domain is less important. An example might simply be a thousand-page manual for an airplane, Dr. Hall says, where readers just want to know “which room to use. “”search” and “turn left or right. “The other type includes literary translations, where the smallest main points count.
The first type has been the most affected by AI. Many employees are now editing translations that have passed through a tool that underpins Google’s translation service. They are paid at a hefty discount according to Word, but there are more jobs available. Lucia Ratikova, a Slovak specialist in legal and structural translations, estimates that structure paintings now account for more than a fraction of all listings on structural sites, up from a tenth a few years ago. More companies, many of which are eager to expand into global markets, are benefiting from declining prices.
If devices can do what humans do more cheaply, employers will turn to computers. But as costs come down, the overall demand for a service is likely to increase, and perhaps enough to offset increased device usage. There is no law whose effect prevails. So far, in the United States, the number of translators has increased, but their real salaries have fallen slightly (see chart 2), probably because the career now requires fewer skills.
Customer service offers more challenging terrain for AI. Companies have been trying to automate it for years. So far, they are the ones that have most commonly bothered visitors. Who doesn’t play with the chatbot to communicate with a genuine human being?The U. S. Visitor Satisfaction IndexThe U. S. workforce has been declining since 2018 and staffing seems to have had enough, too. “Contact center” profit hit a record 38% last year.
But perhaps there’s one consolation: The workforce is becoming more welcoming to the unprofessional. Stanford’s Erik Brynjolfsson, along with Mit’s Danielle Li and Lindsey Raymond, studied deploying an AI assistant to more than 5,000 visiting agents earlier this year. presented real-time suggestions to workers. This increased the productivity of less professional agents by 35%, while that of more professional agents saw little change.
It would be moderate to assume that the effect on salespeople would be very similar to that affecting visitor service employees, but that is not the case. Marc Bernstein of Balto, a company that creates artificial intelligence software for sales groups and call centers, points out that “style points” (i. e. , the air of mystery and the ability to expand a relationship) matter far more in sales than in visitor service, where the vital thing is getting the right answer quickly.
AI could even create sales superstars. Skylar Werneth has been in the industry for 8 years and now works at Nooks, a startup that automates sales. The software analyzes your calls and identifies the most effective tactics. It also helps you call other people. Immediately. Most consumers don’t come to pick up; parallel numbering ensures that Mr. Werneth talks more and hears fewer tones. He estimates that the equipment presented through Nooks will make him 3 times more productive, allowing him to pay much more than before.
What does this mean for labor markets? Sales reps earn bonuses based on the number of consumers they attract above a certain threshold. When productivity increases in a company, bosses tend to raise the threshold. Because not everyone has to reach them, underperforming staff are excluded from the hard work market, as the demand for products does not increase in parallel with business performance, as would be mandatory to justify their retention. The result is a small number of highly productive salespeople. At least, given the highest turnover rate in the industry, moving into this situation may simply mean hiring fewer people, not mass layoffs.
Wow AI
If AI eventually becomes superhuman, as many attendees of the recent UK summit believed possible, all bets are off. Even if AI advances in a less historic way, labor markets will experience as-yet-undiscovered changes. A study by Xiang Hui and Oren Reshef of Washington University in St. Louis and Luofeng Zhou of New York University, published in August, found that earnings generated by writing, reviewing, and editing on Upwork, a freelance platform, fell by 5% after the introduction of Chatgpt last November, relative to roles. less affected by AI. A Balto survey of 400 middle managers found that the percentage of at least some AI increased from 59% in April to 90% in October. Bernstein believes that although “AI today is not able to replace humans [in calling intermediaries]. . . In ten years, maybe five years, she’ll be there. “
The other aspect of AI disruption is the creation of tasks elsewhere. A 2019 model by MIT’s Daron Acemoglu and Boston University’s Pascual Restrepo suggests that the effect of automation is worse for employees when productivity gains are low. little excess wealth to increase the demand for personnel in other sectors of the economy. Our survey of industries on the front lines of AI replacement suggests that the new generation has the potential to lead to much greater efficiency. The picture of inequality remains bleaker. So, it’s better to be a superstar than a newcomer, if only to be safe.
For more expert research on the most important economic, monetary, and market news, sign up for Money Talks, our subscriber-only weekly newsletter.
© 2023 The Economist Limited newspaper. All rights reserved.
From The Economist, published license. Original content can be viewed at https://www. economist. com/finance-and-economys/2023/11/16/what-will-artificial-intelligence-mean-for-your-pay
Advertising
Advertising
Advertising
Ad
Advertising
Advertising
Advertising
Advertising
Advertising
Advertising
Ad