> …can’t people just build their product with openAI or other and plan to move away based on the cost and fit for their circumstances?
> Couldn’t someone say prototype the entire product on some lower-quality LLM and occasionally pass requests to GPT4 to validate behavior?
This seems to reduce the switching costs to almost nothing.
[1] https://news.ycombinator.com/item?id=36626943
tldr; reliable model switching involved multiple 100-line if/else statements, which made our code messy, and debugging in prod pretty hard.
> …can’t people just build their product with openAI or other and plan to move away based on the cost and fit for their circumstances?
> Couldn’t someone say prototype the entire product on some lower-quality LLM and occasionally pass requests to GPT4 to validate behavior?
This seems to reduce the switching costs to almost nothing.
[1] https://news.ycombinator.com/item?id=36626943