In December, staff at the Guild of Writers and Artists of America — a 26-year-old membership organization for contributors — realized something big was afoot.
The latest version of ChatGPT, a “large language model” that mines the internet to answer questions and perform tasks on command, was just released. Its capabilities are astonishing — and exactly what those who make a living generating content such as ad copy and blog posts know well.
“They were terrified,” said Rebecca Matter, the institute’s director. During the holidays, she was busy organizing a webinar discussing the pitfalls and potential of new AI technologies. More than 3,000 people signed up, she said, and the overall message was cautious but reassuring: Writers can use ChatGPT to complete assignments faster and take on higher-level roles in content planning and SEO.
“I do think it’s going to minimize short copy items,” Ms Matt said. “But on the other hand, I think there will be more opportunities in terms of strategy and so on.”
OpenAI’s ChatGPT is the latest in a steady stream of innovations that offer the potential to transform many professions and wipe out others, sometimes in tandem. It is too early to count capable and endangered, or to gauge the overall impact on labor demand and productivity. But it’s clear that AI will impact jobs in a different way than previous waves of technology.
A positive take on tools like ChatGPT is that they can be a complement to, not a substitute for, human labor. However, not all workers are optimistic about the expected impact.
Katie Brown is a grant writer for a small nonprofit in suburban Chicago that addresses domestic violence. She learned in early February that a professional association of grant writers was promoting the use of artificial intelligence software that could automate parts of applications, leaving only humans to polish them before submission.
The platform is called grantable, based on the same technology as ChatGPT, which markets itself to freelancers who charge per app. In her view, this clearly threatens the industry’s opportunities.
“To me, it’s common sense: Which do you think a small nonprofit would choose?” Ms. Brown said. “A full-time salary plus benefits, or someone with AI and you don’t have to pay benefits?”
AI and machine learning have been running in the background of many businesses for years, for example, helping to evaluate a large number of possible decisions and better align supply and demand. Numerous technological advances over the centuries have reduced the need for certain workers—although each time more jobs have been created to offset the jobs lost.
However, ChatGPT is the first company to reach such a broad range of white-collar workers so directly, and so accessible that people can use it in their own work.And it’s improving rapidly, with new version Released this month.according to a poll In a survey conducted by job site ZipRecruiter after ChatGPT was released, 62 percent of job seekers said they were concerned that artificial intelligence would destroy their careers.
“ChatGPT makes it more visible,” says Michael Chui, a partner at the McKinsey Global Institute who studies the impact of automation. “So I think it does start to raise questions about where the timeline might start to accelerate.”
A new generation of chatbots
A brave new world. A new crop of chatbots powered by artificial intelligence has sparked a debate over whether the technology can upend the internet’s economy, rendering today’s powerhouses obsolete and creating the industry’s next giants. Here’s what to know about bots:
This is also the conclusion of the White House Report Regarding the impact of artificial intelligence technology, including ChatGPT. “The main risk of AI to the workforce lies in the general disruption it could cause to workers, whether they find their jobs newly automated or whose job designs have fundamentally changed,” the authors write.
Currently, Guillermo Rubio finds that since he started using ChatGPT to generate ideas for blog posts, write the first draft of a newsletter, create hundreds of subtle variations on inventory ad copy, and call research on a topic, his The job has changed significantly and he might write a white paper.
Since he still charges clients the same, the tool just makes him work less. Still, if the copy rate drops—and it could, as technology improves—he believes he’ll be able to consult on content strategy alongside production.
“I think there’s more reluctance and fear among people, and for good reason,” said Mr. Rubio, who now lives in Orange County, California. “You can look at it negatively, or you can embrace it. I think the biggest takeaway is that you have to be adaptive. You have to be open to embrace it.”
After decades of research, researchers have learned a lot about the impact of automation on the workforce. Economists such as Daron Acemoglu of the Massachusetts Institute of Technology have found that technology has played a major role in widening income inequality since 1980. As unions shrink and systems for training and retraining are hollowed out, workers without a college education have less bargaining power over machines that can do basic tasks.
However, the emergence of ChatGPT three months ago sparked a flurry of research based on the idea that this is no ordinary bot.
group of researchers to analyze Based on models tuned for generative language tools, industries and occupations most exposed to artificial intelligence are shown. Topping the list are university humanities professors, legal service providers, insurance agents and telemarketers. However, exposure alone does not determine whether the technology has the potential to replace workers or merely augment their skills.
MIT Ph.D. students Shakked Noy and Whitney Zhang conducted a randomized controlled trial Experienced professionals in areas such as human relations and marketing. Participants’ tasks typically take 20 to 30 minutes, such as writing press releases and short reports. People who use ChatGPT complete tasks an average of 37% faster than those who don’t – a massive increase in productivity. They also reported a 20 percent increase in job satisfaction.
one third study – using a program developed by Microsoft-owned GitHub – assessed the impact of generative AI on software developers. In a trial conducted by GitHub researchers, developers were given an entry-level task and encouraged to use a program called Copilot, and they completed the task 55 percent faster than those who did it manually.
These productivity gains are unlike almost any that have been observed since the widespread adoption of personal computers.
“It does seem to be doing something fundamentally different,” said David Autor, another MIT economist who advises Ms Zhang and Mr Noe. “Previously, computers were powerful, but they simply and mechanically did what people programmed them to do.” Generative AI, on the other hand, is “adaptive, it learns and it can flexibly solve problems.”
This is evident to Peter Dolkens, a software developer for a company that makes online tools primarily for the sports industry. He has been integrating ChatGPT into his work for tasks such as summarizing blocks of code to help colleagues who may take over the project after him, and proposing solutions to problems he encounters. If the answer wasn’t perfect, he would ask ChatGPT to refine it, or try a different approach.
“It’s the equivalent of a well-read intern,” says Dorkens, who is based in London. “They may not have the experience of knowing how to apply it, but they know all the words, they’ve read all the books, and they’re able to have some success with it.”
There was another takeaway from the preliminary study: ChatGPT and Copilot boosted the least experienced employees the most. If true, this could mitigate the inequality-widening effects of AI more generally.
On the other hand, as the productivity of each worker increases, fewer workers are needed to complete a set of tasks. Whether this will result in fewer jobs in a particular industry depends on the demand for the services provided, and the jobs that may be created in helping to manage and direct AI “hint engineering”, e.g. the skills of those already juggling using ChatGPT are sufficient Long hours can be added to their resume.
With the seemingly insatiable need for software code and the extremely high salaries of developers, it seems unlikely that increased productivity will deter people from entering the field.
However, every profession is different, and Dominic Russo is pretty sure that’s not the case with his: When pharmacy benefits managers and insurance companies refuse to write prescriptions for expensive drugs, they write letters appealing to them. He has been in the job for about seven years, gaining expertise through on-the-job training alone after studying journalism at university.
When ChatGPT came out, he asked it to write an appeal on behalf of a psoriasis patient who wanted the expensive drug Otezla. Turned out pretty good, just needed some editing before committing.
“If you know what to cue the AI with, anyone can do the job,” Mr Russo said. “That’s what really scares me. Why is a pharmacy paying me $70,000 a year when they can license the technology and pay people $12 an hour to run tips?”
To protect himself from this possible future, Mr Russo has been building his sideline: Selling pizza at his house in southern New Jersey, a business he doesn’t think will be disrupted by artificial intelligence.
However.
