What is prompt engineering and how does it work?

Prompt engineering comprises various techniques and methods for optimising prompts for generative AI tools. We’ll explain the definition of prompt engineering, why it’s important, and go over examples and best practices.

Properly formulating prompts for AI tools is imperative if you want to get the most out of language models and produce high-quality results for a variety of use cases, ranging from text and images to video. As artificial intelligence continues to evolve, so has the need for professionals who know how to navigate it most efficiently, which is how the profession of prompt engineer came about.

What is the definition of prompt engineering?

The term ‘prompt engineering’ refers to techniques and methods used to optimise prompts for natural language processing (NLP) and large language models (LLMs) such as GPT-3 or GPT-4, which are based on machine learning. The way a question or instructions are formulated greatly influence the quality and relevance of the answer generated by the artificial intelligence tool. With prompt engineering, the goal is to obtain answers that are better, more precise or more specific.

Prompt engineering for AI models requires creativity and precision as well as an extensive understanding of the language model in question. This is because the choice of words and the order they are placed in can significantly affect the output. Prompts can include natural language text, images, or other types of data.

The same prompt may produce different results on different AI platforms. Therefore, prompt engineering must be adapted for each AI text generator or AI video generator you use.

Why is prompt engineering important for AI tools?

Prompt engineering is essential if you want to achieve better results with generative AI tools and fully harness the potential of language models. For example, a prompt engineer may experiment by posing a question in many different ways to see how it influences the answer. With tools such as ChatGPT, variations in word order and using a modifier once or several times (e.g., ‘very’ or ‘very, very, very’) can significantly affect the results.

For AI image websites, prompt engineering can help fine-tune various features of generated images. These often provide the ability to create AI images in a particular style, perspective, aspect ratio or image resolution. The first prompt is usually just a starting point. The following prompts can be used, for example, to soften or strengthen certain elements and add or remove objects in an image.

Prompt engineering can also help align LLMs and optimise workflows for specific outcomes when developing new tools. There are also other reasons why prompt engineering is important for AI tools:

• Result optimisation: Carefully designed prompt engineering can enable language models to deliver higher quality and more relevant results.
• Efficiency: Well-formulated prompts result in the model delivering the desired information faster, without the need for multiple prompts or iterations.
• Control over output: Clever prompt engineering allows the user to control the way the AI responds, including the length, style and tone of the response.
• Error reduction: Clear and concise prompts help minimise potential biases, misunderstandings or inaccurate answers that a model might give.
• Advanced applications: With proper prompt engineering, AI models can be used for specific tasks or in other areas that they were not originally developed for.
• Experimental insights: Experimenting with different prompts can help gain a deeper understanding of how a particular generative AI works and how it responds to different inputs.

Examples of prompt engineering

Prompts that can be used to create text, images or videos differ significantly from one another. However, for all AI websites, targeted prompt engineering allows users to interact more effectively with the respective AI tool and get answers that meet their expectations.

Prompt example for text generators

Here is an example of targeted prompt engineering for text generators:

1. Specificity

• original prompt: ‘Tell me about trees’.
• improved prompt: ‘Explain the process of photosynthesis in deciduous trees’.

2. Answer formatting

• original prompt: ‘What are the benefits of solar energy?’.
• improved prompt: ‘Name five benefits of solar energy’.

3. Inserting sample answers

• original prompt: ‘Write a sentence about Paris’.
• improved prompt: ‘Write a sentence about Paris in the style of Hemingway’.

4. Length and details

• original prompt: ‘Describe water’.
• improved prompt: ‘Give me a detailed scientific explanation of the molecular structure of water’.

5. Avoiding prejudice

• original prompt: ‘What do you think about cryptocurrencies?’.
• improved prompt: ‘Describe cryptocurrencies neutrally and objectively’.

6. Context

• original prompt: ‘Why do stocks fall?’
• improved prompt: ’Considering economic factors, why might technology stocks fall in a recession?’.

7. Styles or perspectives

• original prompt: ‘Tell me the story of Napoleon’.
• improved prompt: ‘Tell me the story of Napoleon from the perspective of one of his soldiers’.

Prompt examples for image generators

Prompt engineering is not only relevant for language models, but also for models that generate images, such as DALL-E. For image generators, prompts must textually describe what kind of image should be generated:

1. Specificity

• original prompt: ‘Cat’.
• improved prompt: ‘Orange cat sleeping on a blue pillow’.

2. Combination of elements

• original prompt: ‘Buildings and clouds’.
• improved prompt: ‘An old Victorian house resting on floating clouds’.

3. Style and era

• original prompt: ‘Cars’.
• improved prompt: ‘1950s retro-style futuristic cars’.

4. Feelings and atmosphere

• original prompt: ‘Forest’.
• improved prompt: ‘A dark, misty forest bathed in moonlight’.

5. Combination of unusual elements

• original prompt: ‘Table and fruit’.
• improved prompt: ‘A table made of watermelons with a top made of dried banana slices’.

6. Perspective and dimension

• original prompt: ‘Mountains’.
• improved prompt: ‘A huge mountain in the shape of an upside-down tea glass’.

7. Abstraction

• original prompt: ‘Feelings’.
• improved prompt: ‘Joy visualised as a bright explosion of colour’.

Prompt examples for video generators

For video generators, the challenge is to capture not just a single moment or still image, but a dynamic, timed sequence of actions and events. Good prompt engineering helps to precisely specify the action, environment and duration of the video as well as how elements in the video should interact:

1. Action sequence

• original prompt: ‘Cat walking’.
• improved prompt: ‘Orange cat walks slowly past a puddle and then jumps into it’.

2. Environment and mood

• original prompt: ‘Beach scene’.
• improved prompt: ‘A deserted beach at sunset, with gently crashing waves and a flock of birds flying on the horizon’.

3. Temporal development

• original prompt: ‘A growing flower’.
• improved prompt: ‘A rose growing from a bud to a fully bloomed flower in 30 seconds’.

4. Dynamic actions

• original prompt: ‘Sports game’.
• improved prompt: ‘A basketball game in which a player makes a crucial three-point goal in the final seconds of the game’.

5. Combination of elements and transitions

• original prompt: ‘Times of day’.
• improved prompt: ‘A city panorama transitioning from morning to night, with the lights of the city coming on as darkness falls’.

6. Story and narration

• original prompt: ‘A bird flying’.
• improved prompt: ‘A young bird trying to fly for the first time. After a few failed attempts, the bird finally conquers the skies and returns safely to its nest’.

What are best practices for prompt engineering?

With targeted prompt engineering, it’s possible to obtain optimal results from generative AI tools, especially language models. There are some proven best practices that should be taken into account when formulating prompts:

• Be precise: Being clear when wording a prompt helps the AI better understand what you expect it to generate.
• Be specific: Make sure your prompts are specific enough to obtain the type of response you want.
• Experiment: If you don’t get the answer you want right away, try phrasing the question differently or adding more context.
• Format instructions: If you want the answer to be in a specific format (e.g., list, short paragraph, formal language), you should specify this in the prompt.
• Sample responses: Providing sample responses can be helpful as it can give the AI an example of the answer you want and steer it in the right direction.
• Context: Some AIs benefit from being given additional information or more context before the actual question is asked.
• Avoid ambiguity: Avoid unclear or ambiguous wording.
• Limit and direct: If you are concerned that the AI tool may answer in a biased way, or if you want a particular style or perspective, give clear instructions.
• Review: It is important to critically review an AI tool’s responses and ensure they are both accurate and free of unwanted bias.
• Iterative approach: It is often useful to take an iterative approach and refine the question based on the answers received.

What qualifications should a prompt engineer have?

Prompt engineering offers promising opportunities for individuals with a deep understanding of language processing and a creative mindset. As AI and NLP technologies become more prevalent across a wide range of industries, the demand for skilled prompt engineers will continue to grow.

Although there are no requirements in terms of specific education, a degree in a related field can be helpful. Although programming skills are not essential, a degree in computer science or linguistics can make it easier to understand language models and develop prompts. Prompt engineering is primarily about understanding how language works and how to structure it to receive the results you want. The following skills can be helpful in this process:

• Understanding AI and machine learning: It’s important to have a basic understanding of how neural networks work, particularly language models, so you can better understand the mechanisms behind the results.
• Analytical thinking: Analysing results and adjusting prompts based on them requires analytical thinking.
• Communication skills: The ability to articulate clear and concise instructions is essential to prompt engineering.
• Error detection: The ability to detect inaccuracies or errors in an AI model’s responses and make appropriate adjustments.
• Domain-specific knowledge: Depending on which domain you are using it for, specialised domain knowledge may be required to effectively design and evaluate prompts and responses.
• Continuous learning: Artificial intelligence and machine learning are rapidly evolving. Good prompt engineering therefore requires a commitment to continuous learning and a willingness to constantly adapt to new technologies.
• Teamwork: A prompt engineer often has to collaborate with other professionals such as data scientists, software engineers and business analysts.