In the realm of AI, I've been accustomed to providing prompts in a natural language style. This method has invariably been my default approach. Both best practices and my own experiences suggest that giving comprehensive instructions in a conversational manner ensures optimal use of AI tools. It had never occurred to me to leverage AI by merely providing a few keyword prompts.

This concise, keyword-driven approach is instinctive when I conduct a Google search, especially for navigational queries or quick knowledge hunts. For instance, when looking up a Python library documentation or retrieving forgotten command parameters. However, I recently discovered that AI tools can effectively respond to such keyword prompts, reminiscent of the old Google search method.

Unlike searching on Google, where one often sifts through myriad results, AI tools can instantly generate relevant responses. For instance, with platforms like ChatGPT 3.5, short keyword prompts tend to yield quicker results than traditional search engines.


Allow me to illustrate with a recent anecdote. Having not worked with Java for years, I was approached by a friend to assist with his legacy Java system. I intended to implement unit tests for system diagnostics but had forgotten the specifics of using Mockito. My instinct was to turn to Google. However, upon typing "mockito raise exception", I realized I was still in the ChatGPT 3.5 window. To my surprise, ChatGPT 3.5 presented a comprehensive example of how to employ Mockito to craft a unit test for exception handling.


From this experience, I've gleaned several insights:

1. Contextual Keywords: Short keywords can be very effective in procuring quick results from AI tools like ChatGPT. For instance, my "mockito raise exception" query. Keywords should, however, be context-rich. In this case, "Mockito" is known for unit testing, enabling the AI to discern my request for a unit test example.

2. Ideal for FAQ-Type Searches: This approach is exceptionally useful for memory joggers. If you're familiar with a concept but fuzzy on the specifics, keyword prompts can swiftly guide you.

3. Interactive AI Benefits: AI tools trump conventional search engines in the sense that you can quickly ask follow-up questions, facilitating rapid knowledge retrieval. It saves precious time otherwise wasted on filtering through irrelevant Google search results.


However, there are some limitations:

1. Ambiguity: Your keywords must pack sufficient context. Ambiguous or generalized terms that possess multiple meanings can pose a challenge.

2. Accuracy Concerns: The precision of AI tools, including ChatGPT, remains a limitation. It's beneficial to have some foundational knowledge about what you're inquiring about.


In conclusion, the "short keyword prompt" technique is worth experimenting with. If you discover innovative applications for this method or have insights to share, I'd love to hear your thoughts!

The Power of Short Keywords as AI Prompts