In searching library research databases, you need to understand keywords and how to pull them from your research topic in order to get relatable results.
On the other hand, when using AI generators, you do not need to figure out keywords but it is equally important that you incorporate the most relevant information while building the prompt in order to maximize the relevancy of responses from the AI generator (ChatGPT, Gemini, or Copilot).
Unlike traditional research databases, which are run to recognize only keywords, most AI are developed to recognize whole complex sentences and phrases. While research databases and AI generators are designed to retrieve information differently, the outputs for both are still determined by the information input. This means that you will want to be very explicit about how you set up the prompts that you will be drawing upon.
There are a lot of mnemonic (acronyms made of concepts or ideas) frameworks that are floating out there related to prompt engineering (for example, CORE, CARE, ROBOT, PARE, Co-STAR, RISEN, CLEAR, CREATE, etc.) which can all assist with developing the best prompts for AI. The mnemonic deployed for prompt engineering can be aligned to the type of outcome an individual is seeking.
Regardless of which framework you decide to utilize to develop your AI prompt, the most well-developed prompts will incorporate the following five elements:
