Even if ChatGPT doesn’t take your job, it can help another person replace you, says the economist who famously concluded that AI could cut nearly half of US jobs.

  • In 2013, Oxford-based economist Carl Benedict Frey estimated that artificial intelligence could replace almost half of all US jobs.
  • In recent months, new technologies such as ChatGPT have taken the business world by storm.
  • We spoke with Frey about whether ChatGPT is right for Americans to work.

In 2013, about a decade before ChatGPT was first introduced to the public, Oxford University economist Dr. Carl Benedict Frey published a study that scared many Americans.

In an article he co-authored entitled “The Future of Employment,” Frey estimated that 47% of all US jobs could be replaced by automation in “the next decade or two.” The paper has been cited over 5,000 times and featured in countless media outlets, including the 2019 Last Week Tonight with John Oliver.

AI has not yet replaced all of us, and the article is not without criticism, but Frey, who has worked as a consultant for the United Nations, the OECD and several Fortune 500 companies, argues that AI will disrupt many industries in the future.

A decade after the article was published, Insider spoke to Frey about what he did right and wrong, whether ChatGPT is right for all of our jobs, and why, even if it doesn’t replace you, it can help someone else do it. However, not everything is so bad. Frey says that in the long run, AI capture could be good for all of us.

AI may not take your job, but it will have an impact

Frey concluded in 2013 that “low-skill, low-income jobs” are most at risk of being replaced by automation, and he says he still believes this is the case.

“When autonomous vehicles come along, they won’t replace the taxi driver, they will replace him,” he told Insider. He also listed truck drivers, telemarketers, secretaries, cashiers, waiters and photo models as jobs particularly at risk of being replaced.

However, when it comes to ChatGPT and its associated artificial intelligence, Frey says job replacement is “probably the wrong way to go about it,” but workers still have reason to fear for their jobs. Reason: More competition.

Frey points to “creative” industries such as writing, music, and art, as well as the roles of graphic design, advertising, and computer programming, where he says broad access to tools like ChatGPT will enable many more people to produce high-quality work.

So rather than “completely replacing” your job, Frey says, adopting ChatGPT may end up with a different outcome – being replaced by another person. And even if workers manage to keep their jobs, he says the influx of new competitors for their roles could lead to lower wages.

He compares this to the rise of ride-sharing companies like Uber, which has resulted in more drivers, more competition and eventually incumbent drivers facing “a pay cut of about 10% or so.” , according to his research.

“Uber hasn’t reduced the demand for taxi drivers,” he said. “It has, if anything, increased the number of people who make a living driving cars, but it has reduced the size of the earnings of current drivers.”

He says a similar development could take place in industries where ChatGPT is being used most effectively.

“In my opinion, it’s not so much about automation,” he said. “It has more to do with democratization and competition, which could lead to lower wages for people in some of these occupations.”

However, Frey doesn’t think AI will only replace worker roles, citing fashion models and credit analysts as two jobs that could potentially be at risk.

The AI ​​was slower to replace drivers than expected, but also showed it had the potential to do more than Frey thought.

Looking back at the 2013 study, Frey admits that some things he would have done differently had he been able to go back in time. However, he believes the study’s assessments of potential AI job replacements are still “broadly fit for purpose.”

Frey points to one category of jobs in particular that is much slower to automate than he expected a decade ago: drivers.

“I think most of the people we spoke to at the time said it would be 10 to 15 years from now,” Frey said of the widespread adoption of autonomous vehicles. “And 10 years later, most people are still saying it’s going to happen in 10 to 15 years.”

On the other hand, he was skeptical that, for example, waiters and fashion models could potentially be automated, but says new technology has since convinced him that it is possible.

In the distant future, according to Frey, it is likely that artificial intelligence will “replace everyone.” As for his 47% forecast, the paper said it could happen “perhaps in a decade or two”, which many people jumped at.

But Frey says that he and fellow researcher Michael A. Osborne were not going to make an accurate prediction of “the rate of automation,” which he says “cannot be reliably predicted.” Instead, he said, they were more focused on predicting the volume of jobs that could be at high risk of replacement — whenever mainstream AI does materialize.

AI could improve the jobs of the future

Although new technologies have affected many jobs in the past decades, they have not always led to replacement.

Secretaries and bank tellers, for example, do “very different jobs” from decades ago, Frey says, but they haven’t been formally replaced—their roles have simply changed.

“I think there is a somewhat misleading distinction between displacement and jobs being changed and transformed by technology,” he said.

In general, Frey says that technological advances like ChatGPT are good for society. “That’s the reason we’re so much more prosperous today than we were a couple of hundred years ago.”

And for many workers, AI tools can ultimately make their job easier.

“Some of the more boring parts of the job may disappear,” he said. “We can focus more on generating the right ideas, asking the right questions, and more interesting things.”

At least in the short term, Frey says, humans need to maintain an edge over AI in many areas. This is because there are probably some things that ChatGPT cannot learn by looking at data from millions of websites.

“People communicate not only on the Internet,” he said. “We interact in the real world and also use experience there. And often we come up with things from a completely different field that are unlikely to be in the training data set.”

Content Source

News Press Ohio – Latest News:
Columbus Local News || Cleveland Local News || Ohio State News || National News || Money and Economy News || Entertainment News || Tech News || Environment News

Related Articles

Back to top button