Working in artificial intelligence (AI) without having a science, technology, engineering or math degree is possible. And it can pay off big. With the success of these models, the tech industry continues to laud a new profession: AI trainers.
The Anthropic artificial intelligence laboratory is precisely in full investigation, indicates the US site Business Insider. He wants to hire a “fast engineer and librarian”, or in French an “engineer and librarian of proposals and suggestions”.
This role consists of using an artificial intelligence generation tool to improve it. In particular, erasing the stereotypes extracted from the Internet. It mainly focuses on models that create text, such as OpenAI’s ChatGPT or its new version GPT-4.
Whisper to the AIs
Proof of the importance of this function, Anthropic offers between 175,000 and 335,000 dollars annually for the position, as revealed for the first time by the US agency Bloomberg. Applications are open to all, without having to prove any degree in the field.
Here, the specialized laboratory not only wants to find “the man (sic) who whispers in the ear of AIs.” The objective is to build databases that group iterations or sequences of proposals that allow users of a generation tool to find exactly what they are looking for. Anthropic also claims to offer tutorials to learn how to properly formulate your proposals.
The Hays agency recruiter in the United Kingdom, Mark Standen, explains that salaries start at around 40,000 pounds sterling (about 45,500 euros) per year. But some candidates within his database are looking for positions paying between £200,000 and £300,000 (or between €227,800 and €341,700). According to him, “AI-savvy trainers can set their price.”
However, this role should not be a profession of the future. At least that’s what Professor Ethan Mollick of the University of Pennsylvania thinks. “I strongly suspect that ‘suggestion engineering’ won’t be a big business in the long term and that ‘AI coaching’ is not a profession of the future,” he says. A judgment based on the dizzying advancement of build tools as new versions are released.
Source: BFM TV
