1678781421 The boom of artificial intelligence creates a new job

The “boom” of artificial intelligence creates a new job: expert in conversation with the machine

Latest News / By /

Pau Martí Felip, 23 years old and graduate in Audiovisual Communication, started working as a video editor in a digital company two years ago. Over time, he saw his work change: “That’s how I became a fast engineer [ingeniero de peticiones], which is a creative and technological mix of instructing the AI ​​to give you an answer,” he says. The new artificial intelligence applications to create images, text and soon videos or music must receive text requests for what we want. This prompt can range from a simple sentence (“Do some 10-minute stretches for people over 70”) to a complicated eight-line instruction that details, for example, an image’s colors, backgrounds, or techniques.

“People have to understand the machine,” says Martí Felip. “It’s like talking to an animal that most humans can’t communicate with, we’re learning that language,” he adds. A LinkedIn search for “Prompt Engineer” and “Spain” returned only six profiles and one job listing. The offer comes precisely from Martí Felip’s company Raona, which is dedicated to digital transformation in companies and has more than “240 employees”. This unique vacancy for “Request Engineers” in Spain proves its exploratory nature. In February, an offer for a prompt engineer from a major US start-up with an enviable salary went viral.

The job offer is intended to help Martí Felip, who is aware that he is looking for candidates in the deserted country. “Right now, most resumes don’t have AI capabilities. If you ask four questions about ControlNet or something more advanced, you’ll see,” he says. But he is ready to arrange: “Someone who wants to learn and is very creative is definitely worth it.” Although it may seem striking, Raona is now not looking for more computer scientists: “We are full of programmers, but we see that we need more creative people.”

The actual appearance of petition engineers has sparked debate about their feasibility and future. These are difficult prophecies, and even the most skeptical will admit two things: nothing is predictable in this area today, and in the short term (two or three years) such work will be required.

“It’s not like prompt engineering offers seem like Rauch to me,” says Javi López, an investor in startups and AI disseminator (known on the networks as @Javilop). “They seem temporary to me. During a transitional period, many people who have already been trained will need it. But in two or three years it will be trivial and the concept of prompting as such will no longer exist. Even a 5-year-old will be able to ask a machine for things out loud. The difficulty of being a prompt engineer will be zero,” he assures.

an illustrated novel

But López himself has so far received more than €27,000 in funding from 612 donors for an illustrated artificial intelligence novel that includes a guide to getting started and the descriptions (prompts) he uses. That number multiplies López’s original prediction by almost five; There is an interest in exploring this world. Perhaps one day the queries will be simple, but today López’s examples include parts like this string (roughly translated from the original English): “Ultra-realistic, low angle, on a wooden table in a beautiful kitchen, Canon 5D, DSLR, portrait 50mm, DOF, V-Ray rendering, 8k, ray tracing, golden hour lighting, uplight, hard edge lighting.”

López’s skepticism stems from his belief that professionals in any field will know how to handle intelligent machines, a new “engineer” is not needed: “There were also Excel experts and now all accountants are expected to that they know the tools of their work. What remains at the end is not a prompts engineer, but someone who was a designer or artist before. However, the barriers to entry are lowered: it no longer takes as much technique (drawing, lighting, photographing), but much more direction and choice,” he affirms.

Pablo Moreno-Muñoz, a researcher at the Technical University of Denmark, also believes that the models tend to be simplified in at least three ways: “First, training the tools with more data (images, text); 2 hours of engineering to build larger capacity models (number of parameters, size of neural networks); and third, training time and money spent on supercomputers, where AIs find more and more relationships between data that then allow them to generate better results from prompts.”

The key difference in this debate boils down to a simple question: will AIs be able to understand what we mean, like Google search does today, or will it handle the complex requirements of a future video, image, tune, or history still require specialists? It may also be that simple use of these tools comes with something more sophisticated: today you can search on Google or be an SEO expert, and you can use Photoshop or just apply a filter.

Programmer Simon Willison is a proponent of prompt engineering. He believes that in addition to communication skills, they will need something from all of these disciplines: linguistics, understanding how the deep learning that powers these models works, psychology, art history, computer security, and philosophy. “How can we teach a language model the difference between truth and fiction? What is the truth anyway?” Willison wonders.

Silviu Pitis, a researcher at the University of Toronto, also sees a clear future for this profession: “If the models get stronger, we will still need people to interact with them, to teach them to communicate.”

How to train for it

Martí Felip’s specific role is twofold: helping engineers test the model they’re programming, such as ChatGPT, and then using it to achieve optimal results, or helping the customer take advantage of it.

Another key issue is training. Martí Felip misses having learned code at university. But most of his education was autonomous: “I’ve always been interested in AI. I learned with tutorials on YouTube, Twitter accounts and TikTok. In college we should have done a little more code. I know the basics of programming, but I miss something more advanced Python. Although it’s interesting to have the creative part for the design,” he explains.

Pau Martí Felip, at the Raona headquarters.Pau Martí Felip, at the headquarters of Raona MASSIMILIANO MINOCRI

Jessica Gutiérrez, an administrative assistant from Gijón, is another of six people who added “Prompt Engineer” to their LinkedIn bio. She is an administrative assistant but devotes herself to writing for websites. The step was almost mandatory: “Now it takes me a lot less time, if you don’t learn to generate text your copy work will obviously become obsolete,” he says. “I realize that four months since I started researching and getting wet every day, you end up finding an ally in artificial intelligence, although I’ve come a long way to position myself as early as possible,” he admits.

His training was also online: “He was self-taught, trial and error, based on watching videos on YouTube, on Twitch,” he explains. Now he sees the future as an open field: “There is still a lot ahead of us to lay the foundations for job profiles. Participating in the ecosystem is very useful for exploring business and educational opportunities,” he assures, although those around him don’t see it clearly: “People are laughing. They don’t believe that this has a professional future. And they don’t think the tool generates everything I ask for. Between laughter I made the book”. Gutiérrez points to the fact that he created a recipe book in which text and illustrations were all created using artificial intelligence: “I got the recipes in one afternoon. It took me longer to do the illustrations,” he explains. If it works well for him, he will move on with a snap.

And that without code

One of the wonders of this technique is that knowing the code is secondary. A detail that lowers the access barriers. Andrej Karpathy, respected programmer who led Tesla’s AI and just returned to OpenAI (creator of ChatGPT), tweeted: “The hottest programming language is English.”

It is possible, says Ignacio Peis, a researcher at the University of Carlos III, “to classify English as a new programming language because program code routines are generated from a few lines of text. One might think that this is a new level of abstraction. However, the programming relationships between the spoken language and the generated code are not defined, they are not universal. It can be shown that the same text input can generate different codes, since we are talking about probabilistic models here,” he explains.

Therefore, English does not always work perfectly as a programming language. But that doesn’t stop them from taking off as a new way of handling the machine: “In a way, it’s programming,” Pitis sums up. “As a prerequisite for strong AI, computers must be able to communicate with humans through natural language. If that interface becomes strong enough, we can teach them by talking to them,” he adds.

Although English is a possible programming language, machines can beat humans at it too. A recent academic paper tests it with automated queries, trying to improve on a human query, and has success adding to some questions at the end, “Think about it step by step.” In the study, they assert that the best formula would be something like this: “We’re going to solve this step-by-step to make sure we have the right answer.” The race for the perfect request has only just begun, says Gutiérrez, who conducts his tests separately: “You have to be very sure of the request. I write it aside two or three times and if I don’t have all the details I add more and don’t ask the machine until I have it complete.

you can follow THE COUNTRY Technology on Facebook and Twitter or sign up here to receive our weekly newsletter.