Prompt engineering is not what many people think it is. You might have seen ads for prompt engineering courses that promise to teach you how to get surprising answers from AI, complete any task in minutes, and make a lot of money. All this is promised at a “once in a lifetime” cost.
Usually, these offers sound very attractive, but not all courses are worth it. To avoid being disappointed by false promises, the first thing to do is to research the subject you are going to learn beforehand and figure out how it will benefit you.
Let’s break down the concept of prompt engineering and clarify its meaning and role.
Being a key part of generative AI, prompts define how users interact with systems. On its website, OpenAI, the developer of ChatGPT, explains that GPT models are trained to understand natural language and generate texts based on user inputs. The platform mentions:
Prompt engineering is a relatively new profession that emerged with the rise of generative AI apps, and its scope and definition are still being discussed. AI research scientist and OpenAI co-founder Andrej Karpathy previously wrote on X that the title “prompt engineer” is confusing as the role is not only about prompting. He suggested that the term “AI engineer” could be used to describe the job, though it makes the concept too broad. Elon Musk’s definition of the profession is simple. Replying to Karpathy’s post, he said prompt engineering is natural language programming.
In March 2024, social media platforms in China started suspending or removing influencer accounts that advertise AI lessons after many students complained about the quality of the courses. A report by MIT Technology Review says that influencers without a background in AI suddenly started explaining the technology and selling courses about generative AI tools. Buyers of these courses mentioned that lessons lacked actual content, while tutors urged them to purchase the next course for more information.
Take into account that AI models and prompt engineering techniques evolve rapidly. What works today may not work tomorrow, requiring adaptable strategies. However, having an overall idea of how the technology works may help to find solutions faster.
Let’s break down the concept of prompt engineering and clarify its meaning and role.
What Is a Prompt?
If you've used ChatGPT or other AI apps, you know how they work. You input text, and the app responds. You can instruct the bot to provide information, generate text, check grammar, solve logical puzzles, debug code, and more. The message that tells a language model how to respond is called a "prompt." Whether it's a chatbot, AI image generator, or any other AI application, when you type instructions to specify how you want the model to behave, you're creating a prompt.
Being a key part of generative AI, prompts define how users interact with systems. On its website, OpenAI, the developer of ChatGPT, explains that GPT models are trained to understand natural language and generate texts based on user inputs. The platform mentions:
Designing a prompt is essentially how you "program" a model like GPT-4” usually by providing instructions or some examples of how to successfully complete a task.
The quality of AI responses depends greatly on the instructions, not to mention the data it's trained on and the technical capabilities. To improve outcomes, you can experiment with different prompts and use specific strategies. OpenAI recommends providing detailed context, asking the AI to respond in a particular persona, breaking tasks into smaller steps, and trying out other techniques. They note that users are not limited to these suggestions and can explore creative ideas to achieve better results.
Examples of worse vs better prompts. Source: OpenAI Platform
Prompt Engineer ≠ ChatGPT User
Although AI app users write and sometimes optimize prompts, it would be wrong to call them prompt engineers because prompt engineering is a broader field and profession. Besides creating prompts, it involves understanding technical concepts like natural language processing and machine learning algorithms to analyze issues and improve inputs for better results.
Prompt engineering is a relatively new profession that emerged with the rise of generative AI apps, and its scope and definition are still being discussed. AI research scientist and OpenAI co-founder Andrej Karpathy previously wrote on X that the title “prompt engineer” is confusing as the role is not only about prompting. He suggested that the term “AI engineer” could be used to describe the job, though it makes the concept too broad. Elon Musk’s definition of the profession is simple. Replying to Karpathy’s post, he said prompt engineering is natural language programming.
Elon Musk’s definition of prompt engineering. Source: x.com
Magic Prompts and Misleading AI Lessons
The desire to get better results from a bot or learn a trendy profession is why many people consider taking AI courses. In 2024, prompt engineering is one of the most demanded jobs in tech, with a rising number of listings and high salaries. However, the lessons promising magic prompts or exceptional skills in days usually fall short of delivering their promise. Some of them may turn out to be complete scams, taking victims' money and not providing any lessons at all. Other courses may be irrelevant or too basic for learners.
In March 2024, social media platforms in China started suspending or removing influencer accounts that advertise AI lessons after many students complained about the quality of the courses. A report by MIT Technology Review says that influencers without a background in AI suddenly started explaining the technology and selling courses about generative AI tools. Buyers of these courses mentioned that lessons lacked actual content, while tutors urged them to purchase the next course for more information.
Things to Consider if You Plan to Take a Prompt Engineering Course
Learning prompt engineering can be a good idea if you want to improve your communication with AI or start a career. To learn the basics, official app guides, research papers, and different materials available online can be a good start. You can come up with your own prompts, test, and improve to see what works best. Plus, as communication works in both ways, you can ask the bots to guide you on how to input better prompts. That could be fun if you’re determined to find an approach to AI, and team up with it for different tasks.
If you prefer structured learning or aim to start a career as a prompt engineer, there are different courses offered by individual experts and institutions. To choose a legitimate course, review the program’s content, institution reputation, tutor qualifications, and lesson platforms. Don’t make fast decisions based on suspicious ads and take your time for research.
Take into account that AI models and prompt engineering techniques evolve rapidly. What works today may not work tomorrow, requiring adaptable strategies. However, having an overall idea of how the technology works may help to find solutions faster.
Anahit Avetisyan
Author
Web3 writer and crypto HODLer with a keen interest in market trends and recent technologies.
