The Prompt Engineer job title may not sound familiar to most active contributors within the IT industry. But that would be about to change thanks to Artificial Intelligence (AI) systems such as ChatGPT.
And it is that in recent times the boom and euphoria for this type of systems focused on generating content or providing information through a conversational interface has become evident.
Using these types of platforms can generally be described as a chatbot on steroids, with capabilities that simply haven’t been seen before.
Perhaps the most popular example of this so far is found in Artificial Intelligences such as Midjourney, ChatGPT or DALL-E. But the reality is that we would just be seeing the beginning of this new era.
This is the case of ChatLLaMA, the first open source implementation of Meta’s LlaMA language model based on reinforcement learning with human feedback (RLHF), which would be 15 times more powerful than ChatGPT.
That AI from Mark Zuckerberg’s company would be just the spearhead of the second wave that at some point will be freely accessible to any user.
There is a new landscape on the horizon and that means that the labor industry must adapt.
This makes a Prompt Engineer with the new AI platforms like ChatGPT
Some days ago The Washington Post published an interesting piece in which he analyzes exactly how the changes promoted by these systems are reconfiguring the dynamics of the labor market in the IT sector.
Where software engineers and programmers continue to be an instrumental piece, but now there is a certain paradigm shift, where grammar, articulation and good logical structure are fundamental.
The hottest new programming language is English
—Andrej Karpathy (@karpathy) January 24, 2023
For the aforementioned systems to function correctly, the ‘prompts’ are used to carry out the requests entered in text that generate the contents. This is how now companies would be crazy to hire ‘prompt engineers’.
A Prompt Engineer, broadly speaking, is dedicated to “programming in prose” according to the newspaper, with the aim of refining the inputs or text requests that users enter into these systems so that in the end they are more efficient and precise.
In other words, they speak in natural language with these engines with elaborate and precious phrases that meticulously delimit the possible response field of action.
The Twitter thread posted by @karpathy and shared above details in general how this evolution is being lived.