The Future. Companies are paying top dollar for people who can test, tweak, and collect prompts for generative-AI systems like ChatGPT — that can then be used by the company at large. The hope is that having a dedicated prompt engineer will give firms a leg-up in increasing efficiency with AI. But just as businesses rushed to fill crypto and metaverse roles, the new push for AI experts may be another bubble. Only time will tell.
Wanted: humans who can make better AI.
- The hot new job in AI is “prompt engineer,” which are people who continually fine-tune prompts for generative AI tools to get better results.
- They then train companies on their findings so they can use the AI tools better.
- They also save the best answers they can coax out of models as presets so employees at the company can easily access them.
According to Albert Phelbs, a prompt engineer at Accenture-run Mudano, the job is likened to “an AI whisperer.” He writes an average of five prompts a day, interacting with ChatGPT roughly 50 times (which is, in turn, training the AI).
So, if you’re having fun messing around with prompts, just know that a variety of companies are paying a lot for you to do that for them.
- Google-backed startup Anthropic is offering $335,000 for a “Prompt Engineer and Librarian,” while document-review firm Klarity is offering $230,000.
- But outside of tech, places like Boston Children’s Hospital and London law firm Mishcon de Reya are also hiring and offering competitive salaries.
Fascinatingly, the jobs actually make sense for people from very different educational backgrounds. Recruiters say they’re looking for people with either PhDs in machine learning or ethics.