Welcome to Software Development on Codidact!
Will you help us build our independent community of developers helping developers? We're small and trying to grow. We welcome questions about all aspects of software development, from design to code to QA and more. Got questions? Got answers? Got code you'd like someone to review? Please join us.
Post History
No. Software Development is about writing software. Yes, you can stretch almost anything that is an input to software to be "writing software". But arguably the closest analogue in actual Software ...
Answer
#2: Post edited
- No. [Software Development](https://software.codidact.com) is about **writing software**. Yes, you can stretch almost anything that is an input to software to be "writing software". But arguably the closest analogue in actual Software Development to "AI Prompt Engineering" would be *unit tests*. A unit test basically says:
- * I am writing or rewriting software
- * In order to make sure it produces the expected, consistent, results for users, I will create a special set of inputs/configuration parameters/runtime commands and look at the results.
- * The results of a unit test should always be the same, unless the functionality of the software changes. For example, there are many different sort routines you could use to sort a list of numbers, and all of them will, if properly implemented, produce the same result, so you can often rewrite your sorting code to run faster without changing the output.
- Whether or not an AI will produce the same results consistently with the same input is beyond my skill level. But I would argue that *except* for someone writing an AI system who wanted to design various prompts for use as unit tests (or perhaps other testing within the development environment), there is no place for those types of questions here.
- On the other hand, as I understand the [Power Users](https://powerusers.codidact.com) community, AI prompt engineering questions should be on-topic there.
- No. [Software Development](https://software.codidact.com) is about **writing software**. Yes, you can stretch almost anything that is an input to software to be "writing software". But arguably the closest analogue in actual Software Development to "AI Prompt Engineering" would be *unit tests*. A unit test basically says:
- * I am writing or rewriting software
- * In order to make sure it produces the expected, consistent, results for users, I will create a special set of inputs/configuration parameters/runtime commands and look at the results.
- * The results of a unit test should always be the same, unless the functionality of the software changes. For example, there are many different sort routines you could use to sort a list of numbers, and all of them will, if properly implemented, produce the same result, so you can often rewrite your sorting code to run faster without changing the output.
- Whether or not an AI will produce the same results consistently with the same input is beyond my skill level. But I would argue that *except* for someone writing an AI system who wanted to design various prompts for use as unit tests (or perhaps other testing within the development environment), there is no place for those types of questions here.
- On the other hand, as I understand the [Power Users](https://powerusers.codidact.com) community, AI prompt engineering questions should be on-topic there.
- As for considering ChatGPT (or similar) as an actual programming language. *Maybe*. But it is quite a bit of a stretch. There are many applications which indeed have substantive systems inside them that can be used as a programming language - e.g., Microsoft Excel, vim. However, at least at the access level that most people have, ChatGPT is more like a search engine - throw something in and see what comes out.[]()
#1: Initial revision
No. [Software Development](https://software.codidact.com) is about **writing software**. Yes, you can stretch almost anything that is an input to software to be "writing software". But arguably the closest analogue in actual Software Development to "AI Prompt Engineering" would be *unit tests*. A unit test basically says: * I am writing or rewriting software * In order to make sure it produces the expected, consistent, results for users, I will create a special set of inputs/configuration parameters/runtime commands and look at the results. * The results of a unit test should always be the same, unless the functionality of the software changes. For example, there are many different sort routines you could use to sort a list of numbers, and all of them will, if properly implemented, produce the same result, so you can often rewrite your sorting code to run faster without changing the output. Whether or not an AI will produce the same results consistently with the same input is beyond my skill level. But I would argue that *except* for someone writing an AI system who wanted to design various prompts for use as unit tests (or perhaps other testing within the development environment), there is no place for those types of questions here. On the other hand, as I understand the [Power Users](https://powerusers.codidact.com) community, AI prompt engineering questions should be on-topic there.