The democratization of artificial intelligence transforms everyday communicators into AI specialists, marking a shift where language skills and human intuition become more valuable than technical expertise in unlocking AI's potential.
Photo Credit: Rob GrzywinskiOriginally posted on February 20, 2023 on LinkedIn.
The Simplicity of "Hello World"
"Hello world" is a simple computer program that is used to introduce people to programming. The goal is to output the text "Hello, world!" on the screen. The magic incantation needed to output "Hello world" depends on what programming language that you're using. What fascinates me most about Large Language Models (LLMs) is that there is no esoteric language to learn in order to get it to output "Hello world". You simply talk to the LLM just as you might talk to another person. Tell the LLM "Say 'Hello world'." and it responds emphatically "Hello world!".
The Art of Prompting
Knowing how to prompt the LLM to give you a desired response is not always easy or obvious. Do you simply describe the problem (as in zero-shot prompting) or do you provide a set of examples from which the LLM should learn (as in few-shot prompting and in-context learning)? Which approach is best? How many examples should you use? Does the order of the examples matter? Are more instructions better? Is it better to be terse or verbose? A whole bunch of people spent the past few years discovering that the answer is an unsatisfying "it depends". (What's worse is that it's also entirely depends on which model you're using. If you switch model vendors or that vendor comes up with a new model then you have to figure it all out again!)
The Rise of Prompt Engineering
The challenges of constructing a robust prompt has led to the emergence of a new role: the Prompt Engineer. Prompt Engineers are responsible for creating the best possible prompts for LLMs in order to achieve the desired results. This requires a rigorous and systematic approach involving example input data, quality metrics for measuring results, and careful testing and tuning of prompts. As models become more advanced and better understood, the role of the Prompt Engineer is likely to change, but for now it's still a largely unexplored field, with discoveries being made every day. In addition, with further advances in model construction and training, it is expected that crafting a useful prompt will become less of an art and more of a science.
The Democratization of AI
A few years ago Citizen Data Scientists became all of the rage. They are non-experts who use data science tools and techniques to analyze data and gain insights without a formal background in data science. Citizen Data Scientists are typically business analysts, domain experts, or other professionals who work with data but do not have the same level of technical skills as a trained data scientist. With the rapid adoption of ChatGPT (>100M users in under 3 months!!!), I expect Citizen AI to become the next new thing.Unlike most technology-based tools, LLMs make it easy for anyone to make remarkable discoveries regardless of their math or science background. In fact, it is those with strong language and people skills which may be better suited to take advantage of the possibilities that LLMs offer. If you are an effective communicator, the role of a Prompt Engineer may be the perfect avenue for you to explore.
New Roles for a New Era
But I think that "Prompt Engineer" may be the wrong name, especially if we want to draw in people from non-Engineering disciplines. "Prompt Professional", "Knowledge Navigator" and "Robot Whisperer" all have a nice ring to them:
A Prompt Professional is a business person who has taken on the role of Prompt Engineer to help their business be more effective, make their job easier, and drive more revenue. They are excited about the possibilities of what Prompt Engineering can do for their business, are constantly looking for ways to improve the effectiveness of their prompts, and they are always trying to learn more about how to use Prompts to achieve their goals.
A Knowledge Navigator specializes in helping people find information. They use LLMs to quickly and accurately answer questions, provide insights, and provide recommendations. They understand the intricacies of LLMs and the best ways to prompt them to get the desired answer. They work with their customers to understand their goals, their data, and how to best structure their prompts to get the best results.
A Robot Whisperer is able to connect with LLMs and understand how to get it to respond in the desired way. They are able to phrase a prompt to get the desired behavior, and how to troubleshoot when things don't go as planned.
Which role do you think best suits you?(896 tokens)
Large Language Models represent the fusion of simple prediction with emergent complexity, where teaching a machine to guess "what comes next" somehow births understanding. Like a child learning language through pattern recognition, these neural networks transform basic word prediction into sophisticated language processing, challenging our assumptions about intelligence and computation.13 January 2025
The "Yeah, But Can It…?" mindset reveals how resistance to AI often masks deeper questions about human value and adaptation. Moving beyond the reflexive hunt for AI's limitations opens paths to meaningful collaboration and innovation, transforming skepticism into opportunity.11 January 2025
