Prompt engineering (PE) is an emerging discipline within the field of Artificial Intelligence (AI). With the arrival of ChatGPT, Midjourney and other conversational / generative AI systems, one needs to interact with them in a specific way to get the desired response. Designing the input to the system so you get an optimal response can be difficult. This is where PE comes into the picture.
To illustrate what PE is, we can use a normal internet search as an example. Using a search engine to find exactly what you are looking for requires the user to act in a specific way. If you want to limit the search result to a certain country or within a specific period in time you need to refine your search query. Over the years, the search systems have improved so you can converse with them with using natural language but you still need to be careful how you converse with the system to get the desired result.
Similarly with the new AI systems, you need to adapt your input so it gives the best result. This is what Prompt Engineering is about. How can I create the right input into the system, “the prompt”, so it generates the best possible response? To the average user the new AI systems can feel very mature, but if you require something very specific to be generated, the prompt needs to be very specific. A prompt engineer will have in-depth knowledge on how to utilize the system in such a way that you get the desired response.
Overall, prompt engineering represents an exciting new direction for NLP (natural language processing) research and development, and has the potential to significantly improve the performance and versatility of AI models in a range of applications.
Read the article How AI and language models can change the energy industry >
At Cegal, we have many consultants who have extensive experience of a range of ML and AI solutions.
Here are some examples of assignments we have carried out for our customers: