Tech

Writers Are Becoming ‘AI Prompt Engineers,’ a Job Which May or May Not Exist

Freelancers say they’re quitting to become ChatGPT whisperers. But is it a legitimate career path, or just another short-lived gold rush?
GettyImages-1065679054
Image: Getty Images

“I literally lost my biggest and best client to ChatGPT today,” a Reddit user going by Ashamed_Apricot6626 posted on the freelance writers subreddit, claiming that the text-generating AI tool had caused them to stop writing and sign up to be a driver for Doordash. The post went massively viral on Reddit and Twitter, and has led to doomsayers suggesting the end of the content writing profession.

Advertisement

The user is just one of many freelance writers on Reddit openly worrying about the future of their profession. Amidst the ongoing hype over content-generating AI models, posts have begun to crop up across the site as writers fret about their careers when AI tools can generate paragraphs of passable text in mere minutes at a fraction of the cost. 

For some copywriters, content marketers, and freelance writers, the solution has been to reinvent themselves as “prompt engineers,” a term that emerged in response to the rise of large language models like ChatGPT. But is it a legitimate career path, or just another short-lived trend riding the AI hype?

The idea of a “prompt engineer” came in response to AI text and image generators, which people noticed would produce higher-quality content when users give specifically-phrased prompts, as opposed to just asking naturally. The title describes people who are supposedly skilled at selecting the right phrasing so that AI tools generate the most accurate and relevant responses. Some job offerings now claim to pay as much as $335,000 and don’t require that applicants have a tech background—several times higher than the average freelance copywriter salary in the United States, which is $55,238.

Advertisement

The trend has caused some to believe that prompt engineers will soon be more in demand than traditional writers. When freelance writer and Reddit user going by the handle “hornylittlegrandpa” added “AI prompt engineer” to their Upwork profile, they claim they got an interview in under an hour. “It was a real eye opener. Like it or not, ‘prompt engineer’ is currently a gig with a lot of opportunity and very little in the way of rules or regulations,” they wrote in a post. “It’s also presently a pretty cheap skillset to develop.” 

A career site called Prompt Talent was created two weeks ago and is dedicated to helping people find prompt engineering jobs. Companies including Huawei Technologies, the China-based telecommunications company and largest 5G network provider in the world, have posted listings on the site. 

“These prompt engineering jobs get filled extremely quickly. In the 2 weeks the platform has been live, 30% of job listings that were on Prompt Talent have been closed,” the site’s anonymous creator told Motherboard and shared a screenshot of the site’s dashboard. 

Motherboard was unable to verify the veracity of these claims, and its creator declined to identify themself when reached for comment.

Advertisement

“This is a wakeup call to lawyers, accountants, consultants etc (read this to stay ahead)”, Greg Isenberg, co-founder of the product design agency Late Checkout, tweeted, after claiming that his friend just lost her copywriting job. This was followed by Isenberg’s advice for people to reinvent their skills to make them AI-proof, redirect their careers in a new direction, and embrace AI as a collaborative tool. 

These types of posts across social media and news sites have contributed to the AI panic, causing some writers to say they are shifting their careers to involve AI and looking for new opportunities as prompt engineers. Motherboard spoke with a number of people who work as professional writers who are worried that their careers are no longer viable without incorporating AI, and are embracing this shift. 

When Peter Leshaw saw a Bloomberg article claiming that prompt engineer is the newest in-demand, high-paying job, he immediately switched his title from marketing strategist to prompt engineer. He told Motherboard that when reading about the title’s responsibilities, a lot of them were skills he had already been using in his everyday work. As a marketing strategist for over a decade, he wrote his own strategy and worked with content writers to create relevant marketing copy, but since January, he has been using GPT-4 to generate prompts for his digital marketing work, including SEO and content writing. 

Advertisement

“If the copy is unique, if the copy is authentic, if it serves its purpose and its place on the Internet that no one else has, why would you not use it?” Leshaw told Motherboard. Because he is able to refine and input context-aware prompts that reflect SEO strategy and target specific audiences as a prompt engineer, he says that his content is original and meets his standards. 

Leshaw said that using AI helps him be a lot more efficient in his workflow, and therefore gives him more time to maximize the number of projects he’s able to work on. “Something that would have taken me eight hours to do the research, formulate the bullet points, I can just do literally in one hour and then I ask ChatGPT to write a blog post about it, promoting it, and then I can put everything on my site and now I got a new course that I can even train,” he said. 

Like Leshaw, Anna Bernstein was drawn to the prompt engineer role by its promise of growth and consistency, which freelance writing often lacks. Bernstein, a prompt engineer at a startup called Copy.ai, has had more experience than most of her peers, having been in the role for over a year and a half now and started on GPT-3, as opposed to the latest models of GPT-3.5 and 4. Bernstein was once a freelance historical research assistant and copywriter but felt that she wasn’t built for the stress of freelancing and having to regularly find new gigs. She told Motherboard that starting her job as a prompt engineer has brought a lot more financial stability. 

Advertisement

Bernstein echoed that most prompt engineering for writing doesn’t require a technical background, but instead, requires the skill set of a strong writer. “Good prompt engineering mainly requires an obsessive relationship to language. It requires both writerly intuition and an intensely analytical approach to what you’re doing, applied at the same time. It also requires creativity—you have to be able to make leaps and think outside the box, especially when it comes to developing new strategies and forms of prompts,” Bernstein told Motherboard. “At the same time, you also need to be the kind of person who is willing to obsessively try variations of the same thing over and over again to see if you can get it right.” 

Bernstein said the job allows her to learn while she goes, formulating new strategies rather than relying on a static strategy. “What’s far more useful to me and my employers is the actual research and development aspect—being able to innovate, roll with the punches of new models, come up with new solutions and mechanics, and unlock new capabilities of this technology,” she added. 

Todd Norem, a freelance creative director and copywriter, was drawn to AI prompt engineering after seeing that many people with a background similar to his were exploring the same transition.

Similarly to Bernstein, he sees being a prompt engineer as more about becoming an expert at adapting to the latest large language model and maximizing its potential, than having a static set of hard skills. “I’m at my best when I’m learning. And I can’t think of a more exciting thing to be involved with. I see AI as a massive tool that is constantly changing. I want to stay up to date on it,” he told Motherboard. I think you have to be a deeply curious person to succeed at this. For me, it’s fascinating that this technology exists, and to be successful you need to understand how it learns and processes prompts.” 

Advertisement

Norem thinks that being able to master and incorporate prompt engineering into his work as a writer will allow him to stand out more and bring more security to his job. “The advertising industry is notorious for layoffs and short stints at different companies. I do think there will be more stability because you’re entering a new frontier. You’re forced to learn and keep up with the technology,” he said. “Those that embrace AI will succeed and income will follow.”

John Benedict, a freelance creative director and copywriter, said he added “Prompt Engineer” to his LinkedIn immediately after reading the same article as Lewshaw, whose headline claimed that this role could make up to $335,000 a year. 

Benedict emphasized that the role still very much requires human creativity and says that his past skills and experience as a writer are still valued and applied while using AI, rather than replaced. “Prompt Engineering, far as I can tell, stripped of the API specifics and whatever best practices have evolved into by today, is creative direction, but where your junior creative teams are artificial,” he said. “So human creatives, while expensive in time and money, are still exclusively in possession of the only skill that matters: knowing good when you see it.” 

Not every writer, however, sees the value of attaching the title of prompt engineer to their resumés. David Morelo, a content writer and copywriter, thinks that AI will undoubtedly decrease value the demand for traditional human content writing. However, Morelo sees AI as becoming ubiquitous, as part of many applications writers already use, and that everyone should be expected to be using AI as a tool to remain competitive. 

Advertisement

“Writers don’t advertise their ability to use a computer—that’s just given. Once AI is included in Word and every other application under the sun, its use will become accepted and expected,” Morelo said. “Content writers who decide to call themselves ‘prompt engineers’ will stand out in a negative way because it will be expected that their content is 100% AI-generated.” 

Morelo, who mainly ghostwrites content, wants to start prioritizing bylined work and putting more weight behind his name in order to stand out during this AI writing boom. “The idea is that the SEO-focused clients I typically write for will be increasingly looking for established writers with relevant industry experience and expertise to meet Google’s new E-E-A-T requirements,” he said, referring to the company’s method of evaluating its search ranking system. “Besides that, I’m now putting more emphasis on my web dev and content strategy skills, trying to demonstrate my ability to handle content production end to end.” 

What makes many writers even more fearful is that prompt engineering is likely just a temporary solution. Morelo has already decided that if he can no longer write for his clients, he would teach English as a second language, which is what he got his undergraduate degree in. 

AI researchers are also pushing back against the current hype surrounding this role, with many believing that it isn’t something they can see lasting in the long term. Ethan Mollick, a professor at Wharton School of the University of Pennsylvania, tweeted that as AI gets more advanced, prompting them to produce high-quality content will become easier. 

“I definitely think there is too much hype around the extent to which AI will replace humans—in no way will prompt engineers replace human writers or poets, because language models don't have all of the creativity and depth that people have, Sasha Luccioni, an AI researcher at HuggingFace told Motherboard.

And while it’s true that AI and automation are already causing major labor shifts, Luccioni notes that it’s in the interests of the companies making these models for people to believe that they will completely replace entire sections of the labor market.

“I would say that artists and authors can learn to use these tools to help them on the boring bits of their work—describing scenes or generating descriptions—but they shouldn't see them as a threat to their talent,” Luccioni added. “And of course the people making these models will say the opposite (just look at Open AI's recent work), but I consider that more as marketing in order to get people to use their product, rather than true estimates of the labor impact of AI models.”