[FEATURE] Embeddings vs information in context - cost difference

[karolklp]

Is your feature request related to a problem? Please describe.
examples/Question_answering_using_embeddings.ipynb should include info about token usage of embeddings vs putting additional information into the context window of the model - aligned with how it's linked from https://platform.openai.com/docs/guides/embeddings "we explore the tradeoff". And it's in the code

Describe the solution you'd like
Table comparing cost of both approaches