How Does One Qualify as an AI Whisperer, and Is It the Right Fit for Your Skill Set?

Quick Links

Key Takeaways

AI Whisperers, or prompt engineers, utilize their understanding of generative AI to yield desired outputs by inputting specific prompts. This role, an intersection of technology, psychology, and art, is gaining relevance as the adoption of AI technologies like LLMs grows. Despite the evolution of AI, human touchpoints in managing and directing AI will remain valuable.

New jobs are being invented all the time, and thanks to the explosion of generative AI, we now have yet another potential new job title on the market: the AI whisperer. Is this another gimmick or your next career?

What Is an AI Whisperer?

While you may not have heard the term “AI whisperer,” you’ve might have heard of the term “prompt engineer,” especially in the context of generative AI art systems such as MidJourney and Stable Diffusion .

A prompt engineer is someone who figures out what the right input is into the black box of generative AI to get the desired output, but if you put “AI Whisperer” into a search engine you’ll see that this newer term for prompt engineer is already in common usage, as the whole area of generative AI takes off at a geometric rate.

To me, the term “AI whisperer” certainly is a better reflection of what this job actually entails. It’s not an exact science by any means. An AI whisperer is shaping up to be a combination of technology enthusiast, psychologist, and (dare I say it) artist.

Where the average person might struggle with AI interactions and feel like they just can’t get a generative AI engine to do what they want it to do, the whisperer can figure out the magic words that spur the AI into action, spitting out exactly what was asked. It’s like watching people who can look at a Rubik’s cube for a few seconds and then solve it instantly.

Click here

Why Do We Need AI Whisperers?

The big question that’s probably at the front of your mind is why we need these “AI whisperers” in the first place. After all, you can put any two people in front of, say, Microsoft Word, and they’ll get the same results just by following the instructions in the manual. You don’t have to cajole Word or Photoshop to cooperate with you—though certainly, having polished skills with either tool is a bonus.

Generative AI engines are completely different. Large Language Models (LLMs) such as ChatGPT and Google’s Bard are examples of machine learning. They have been trained rather than programmed, and while even a hand-coded program can have unexpected behavior, that’s nothing compared to an LLM.

These artificial neural nets are so vast and complex that you can never predict exactly what will come out the other end when you run a prompt through them. On top of this, they have an element of randomness built in by design. This is why giving the same prompt to ChatGPT multiple times never results in exactly the same results.

What an AI whisperer does is intuit what prompts are needed to get a specific result. Because the inputs to these LLMs are in the form of natural language, people who have a knack for language, logic, and communications suddenly have a new role to play as the world rushes to adopt this technology.

Think of the generative AI like the mythical Djinn. You get to make a wish, and the Djinn can give you virtually anything you want, but you have to phrase your wish perfectly, because it will give you exactly what you asked for, not necessarily what you wanted. The more complex and nuanced the wish, the more complex and nuanced your request must be.

Is This Really a Job With a Future?

Most jobs go away eventually. In theory, a combination of AI and advances in robotics replace virtually every type of physical and mental labor over the next century—or at least completely transform them. That’s maybe a little too philosophical an answer, but it helps give us some perspective here.

Related: What Is ChatGPT Plus?

In more practical terms, as these tools advance and become more self-directed, self-correcting, and (gasp) perhaps even self-aware , the need for a human interpreter might become less common.

Then again, perhaps human nature means we’ll always want a human in the loop to babysit our automated systems to one degree or another.

Famously, in Isaac Asimov’s Robot novels, Dr. Susan Calvin is a robopsychologist—a person who specializes in the psychology of positronic robots. Because the brains of Asimov’s robots are so complex (and despite his “three laws “), they can act in unpredictable ways. Sometimes, they have to be persuaded or talked into doing something or to stop doing something.

AI whisperers might be an early iteration of this type of role. Someone who doesn’t think like a programmer, because the “program” doesn’t act like one anymore. AI whisperers might have the knack to get an AI to circumvent its own safety parameters, something we saw people do with glee and lead to some very NSFW behavior from certain AI .

On the less black hat side of things, the more legitimate form of this new breed of computer operator is more likely to be getting these AI to do the positive work we want them to do. AI whisperers will work around their occasional random tantrums and hallucinations , and come up with reliable operating procedures. They’ll likely also play a role in quality control, making sure that the AI’s output matches what their employer wants.

No matter how smart our tools become, it’s likely we’ll always want that AI to work in tandem with a human being. From that perspective, AI whisperers and whatever job eventually follows will probably be around for some time.

TechiDaily

How Does One Qualify as an AI Whisperer, and Is It the Right Fit for Your Skill Set?