Generative artificial intelligence (AI) is what powers tools such as ChatGPT, Google Bard, and Midjourney, which are trained on human-created content and can generate images or text in response to prompts. It promises to automate a lot of tasks we now do ourselves.
That sounds like a boon, but it has also caused people to fear for their jobs. Indeed, Goldman Sachs issued a report(Opens in a new window) on the effects of AI on economic growth that said, “If generative AI delivers on its promised capabilities, the labor market could face significant disruption,” estimating that 300 million jobs could be impacted.
Impacted doesn’t necessarily mean terminated, though. The report goes on to say, “Worker displacement from automation has historically been offset by creation of new jobs” and that 63% of current jobs would be complemented by AI—especially in fields such as customer service(Opens in a new window). The Wall Street Journal reports(Opens in a new window) that the jobs for which AI can do at least 50% of the scut work include accountant, writer, mathematician, PR specialist, and blockchain engineer. Short-order cooks and motorcycle mechanics, your jobs are safe.
As with any emerging tech, generative AI will also create new roles for people to fill—some that even weeks ago, no one had ever heard of. Here are just some of the job descriptions you’ll be seeing more of in the coming years.
1. Prompt Engineer
(Credit: Google Bard/PCMag)
Generative AI uses natural language processing (NLP), which lets you ask for something using normal words and syntax. You don’t have to be a programmer, but that doesn’t mean you can’t perfect what you say.
Prompt is the term for what you ask of AI. A prompt engineer is someone who can compose an effective string of text that not only helps with the training of the AI but also delivers exactly the results needed and intended. A prompt engineer (PE) has to master the tricks that work on specific generative AIs—what you say to Bard may not be the right thing to say to ChatGPT, for example. Prompt engineering is as much an art as it is a science. The CEO of OpenAI (creator of ChatGPT) says great prompt writing “is an amazingly high-leverage skill.”
Current advertised pay rates for a PE can range from $250,000 to $335,000 per year, and the job doesn’t necessarily require a computer-science degree.
Tech companies won’t be the only places that require successful prompts. Expect to see PE jobs for any business that needs AI automation, including law firms, customer support, and, well, publishing(Opens in a new window). At online marketplaces such as PromptBase(Opens in a new window), you can purchase a prompt or sell your own. This is one area of AI where the human touch will be a necessity.
2. AI Trainer
As companies implement generative-AI systems that are meant to interact with the outside world—think AI chatbots doing customer service—AI must learn a few things about the business it’s a part of. That’s what an AI trainer does.
One job post found online as of this writing, for an insurance company that’s implementing an AI platform, specifically says that the applicant will not be speaking to anyone on the phone directly; they will assist the AI platform while the AI conducts calls. Perfecting things that aren’t human so they’ll sound more human is what AI training is all about.
3. AI Auditor
It’s not enough to get cool or interesting results, the output from an AI should also be accurate and unbiased. We have heard plenty about the bias inherent in AI, and the companies powering today’s AI admit it can still happen. The job of the AI auditor is to make sure those problems are weeded out as much as possible(Opens in a new window). In some places, that is, or will be, the law—New York City, for example, has a law coming in July 2023 requiring bias audits of tech used for AI hiring(Opens in a new window).
Recommended by Our Editors
4. AI Ethicists/Ethics Experts
Bing Chat AI (powered by GPT-4) defines this role as the person “responsible for ensuring that AI is used in a safe and ethical manner.” A recent job posting(Opens in a new window) for Indeed.com says its AI ethicist (“staff data scientist”) will be “helping all people get jobs by reducing bias and increasing fairness in our algorithmic systems.”
We asked Google’s Bard AI about the job. It said, “AI ethics experts can command high salaries, as they are in high demand by companies that want to ensure that their use of generative AI is responsible and ethical.” The high demand isn’t a given just yet, but the money isn’t bad—the job mentioned above and a few others we found online topped out at $160,00 per year.
5. Machine Managers
AI isn’t going to be all about people typing natural language and getting back text (or pictures or video). Actual machines are going to have to learn to work with AI instructions. That’s where machine managers will be needed—to oversee AI-operated hardware and systems.
Because of the sophisticated requirements, these job descriptions (which are frequently placed by manufacturing industries) tend to ask for actual computer-science degrees along with commensurate experience in the field in which the AI is being used. The pay can fluctuate, with top salaries of $109K to $251K a year. That seems a little low, though, if a pipsqueak prompt engineer commands a higher salary and doesn’t have an advanced degree. But this kind of hiring is also, apparently, as much an art as a science.
Seeing a lot of overlap in these jobs? You’re right. It’s Wild West time for generative AI, and no one knows which related jobs will stick around. But getting in on the ground floor with a job like any of the above is a smart move. And plenty of existing job titles—old-fashioned ones such as programmer—will still be needed, even as AI proliferates.
Get Our Best Stories!
Sign up for What’s New Now to get our top stories delivered to your inbox every morning.