It is the wave of the artificial intelligence (AI) world that others are trying to imitate.
In the four months since its launch on November 30, ChatGPT has proven its ability to handle a variety of tasks, from breaking the bar and taking the US medical licensure exam to writing emails and songs, building apps, and more. The fact that it is freely available for public use has opened up a range of possibilities previously thought beyond the reach of artificial intelligence, although the program’s creators have been criticised for not being transparent about the programming used to run it.
Developed by Microsoft-backed OpenAI, ChatGPT became the world’s fastest-growing consumer app two months after launch, with more than 100 million users in January.
This early success led Microsoft to integrate its Bing search engine and Edge browser with technology using ChatGPT to improve the user experience. Last week, Google released a similar AI app known as Bard after announcing a preview of the platform in February.
Also in mid-March, the Chinese tech giant Baidu announced its answer to the American app—a platform called Ernie. Both Bard and Ernie stumble early as the race for AI gear heats up. Meanwhile, OpenAI has released GPT-4, which is an updated version of the technology behind ChatGPT. The new platform can analyze images and larger texts (up to 25,000 words), create a website from a hand-drawn sketch, and recreate games in seconds. Companies are now rushing to launch products based entirely on ChatGPT in areas ranging from customer service to financial analytics.
With the International Labour Organisation already predicting that 208 million people will be unemployed in 2023, will this new wave of AI significantly increase unemployment? What jobs can these tools replace? What is the future of work?
Short answer: ChatGPT and its competing AI models have the potential to dramatically disrupt the labour market and replace routine jobs in some industries. But in general, technology can improve productivity and complement workers rather than causing unemployment, experts told the sources. What makes ChatGPT and other similar platforms fundamentally different from previous generations of AI is GPT, which stands for generative pre-trained transformer. Simply put, these tools use a technique called deep learning to generate and analyse text, answer questions, and perform other tasks related to language and speech in a way that mimics humans better than ever before.
“The impact of generative artificial intelligence on the labour market is really important,” Laura Nurski, fellow and head of the Future of Work team at the Bruegel think tank in Brussels, told sources. “
It is widely used as an advanced search engine for gathering information, rough formatting of texts, and creating texts in a certain writing style. This is true for many jobs and almost all fields.
However, the nature of these AI platforms and their focus on language translation tasks means that some professions will be more affected than others, Nurski said.
A new study by researchers at Princeton, Pennsylvania, and New York University found that telemarketers and teachers may be most affected.
The researchers used a benchmark known as AI Occupational Exposure to assess how disruptive services like ChatGPT might be to different occupations. They concluded that even within the education sector, the impact would be greatest on post-secondary teachers of languages and literature, history, law, philosophy, religion, sociology, political science, and psychology.
However, this disruption does not necessarily mean that AI will take millions of teaching jobs, the researchers behind the study and other analysts note. Instead, these new tools could help teachers with some of their tasks, from detecting cheating and plagiarism to developing teaching materials. “It certainly automates some tasks, but that doesn’t mean AI can do all your work,” Nurski said. At the same time, the limitations of AI may limit its ability to meaningfully replace humans. Along with its successes, ChatGPT has had its fair share of mistakes, as acknowledged by its creators, who believe the technology is still “flawed and limited.” It failed, for example, in basic mathematical calculations and logic. Of course, some jobs may become redundant.
“Research on the displacement effect has shown that it is easier for artificial intelligence to replace routine jobs such as translators or telephone operators,” said Georgios Petropoulos, a researcher at the Massachusetts Institute of Technology (MIT) and Stanford University who examines the impact of new technologies on the labour market.
But AI also has “the potential to create jobs,” Nurski said. Indeed, the World Economic Forum noted in October 2020 that while AI is likely to eliminate 85 million jobs worldwide by 2025, it will also create 97 million new jobs, from big data and machine learning to security and digital marketing.
“It’s safe to say it’s going to change the way we work,” Nurski said.
Makes work more efficient
According to economists and technology analysts, this change can be seen in part as an improvement in productivity. There is early evidence of this.A new study by two MIT researchers, which has not yet been peer-reviewed, found that ChatGPT significantly increased the productivity of college professionals performing mid-level professional writing tasks.
The researchers asked 444 writers, consultants and HR professionals to write press releases, summaries, analysis plans and sensitive emails. Half of them used ChatGPT.
The study showed that especially low-skilled workers benefited from this tool, helping to reduce the time spent on their work and reducing the quality gap between their work and skilled workers. For Petropoulos, this is no surprise. His work has shown that even during previous industrial revolutions, employment may have dominated in the short term, but in the long term, as markets adapt to automation disruptions, productivity gains actually create the basis for new job opportunities.
For example, the appearance of the car reduced the importance of those whose jobs depended on horses. But within a few years, the success of the car industry meant a demand for more and more cars – and new jobs were created as a result.
In addition to education, the study, which looked at jobs most likely to be affected by tools like ChatGPT, identified legal services and securities, commodities and investments as sectors that could be disrupted. Morgan Stanley, one of the world’s largest investment firms, is already developing its own AI tool based on ChatGPT. But as with teachers, this impact on productivity could simply mean that AI is taking over some of the more mundane tasks in these occupations, rather than replacing work wholesale, experts said.
They say that artificial intelligence cannot replace non-routine and necessary jobs, such as a lawyer who has to argue in court, or the sensitive role of a pharmacist who has to sell prescription drugs.
“Even though you can use a tool like ChatGPT to help you build a legal case, a lawyer still has to go to court to defend their client,” Petropoulos said.
Ultimately, the economy — and individual businesses — need a balance that combines human and artificial intelligence work, he said. As he noted, even the founder of Tesla and SpaceX Elon Musk, who does not readily accept mistakes, “recognized a few years ago that Tesla’s over-automation was a mistake and that human work must complement the technology.” And there are professions where AI may struggle to make an impact.
No pulls
The study, which looked at the industries most likely to be affected by ChatGPT, found that physically demanding jobs such as textile workers, masons and carpenters remain largely unchanged.
Most experts believe that even highly skilled roles are unlikely to see as much disruption as mid-level jobs – where AI can at least somewhat mimic knowledge.
“Machine learning has not yet reached a point where it can necessarily replace jobs that require highly skilled professionals,” Guy Michaels, associate professor of economics at the London School of Economics, told Al Jazeera. “We are witnessing a turn of workers in the middle phase of knowledge dissemination.”
Brugelsin Nurski said workers whose industries are affected can also upskill, including understanding and adapting to tools like ChatGPT. For all its promise—and the threat of disruption—the ultimate success or failure of AI may not be determined by technology, but ironically by people. Ultimately, the impact of language modeling tools like ChatGPT depends on how consumers respond to the services offered by those platforms, not the people, Michaels said. “Consumers may not find real value in machine-generated content,” he said.
In 2020, a team of American and Chinese researchers surveyed 670 online shoppers about their experiences with AI customer service. Most said they liked that the AI was able to answer their questions faster, 24/7 and more objectively. But most also thought people were more likely to give them more accurate and comprehensive information.
Another study, also conducted in 2020, surveyed Australian hotel customers. It concluded that customers preferred to interact with people.
In other words, while AI-driven change is upon us, it’s unclear what it will look like.
“We’ve seen other technological breakthroughs in the past, from the invention of the motor to the introduction of computers in the workplace. We are still working, the nature of our work has just changed, Nurski said. “Although technology can be disruptive, we do not see the end of the work.”
Get latest news and insights of Global technology industry here