You can use a Large Language Model (LLM) such as ChatGPT as an interface between humans and software.

The LLM can understand the unstructured output of humans and turn it into the input for a computer program. This way, computer programs can grasp the world beyond numbers and structured data.

This naturally opens up a lot of possibilities. Things that were previously not computable, now become something that software can work with.

A simple example: LLMs can be used to detect certain triggers in a conversation that will call a specific function in an application. This could happen for instance when a user in a conversation with a chatbot asks for help in a certain area. Or it could be used to end a chatbot conversation once a certrain kind of information, such as a telephone number or email address have been extracted from the user.

In this way, LLMs enable computation in previously non-computable domains of the world. Pretty cool.

To learn more about how to implement this in your software, read about Function Calling in Openai-Powered applications.