* On 13 June, OpenAI introduced their new function calling API<sup>[1]</sup> * These functions are inserted into the prompt in a non-trivial format<sup>[2]</sup> * Neither `openai-python` nor `tiktoken` currently provide a function to calculate the expected token consumption of a set of function definitions [1]: https://openai.com/blog/function-calling-and-other-api-updates [2]: https://community.openai.com/t/how-to-calculate-the-tokens-when-using-function-call/266573/6
