1. For an AI engineer, you can already build such a system yourself quite trivially by fine-tuning a lightweight inference model, deploying it behind a FastAPI endpoint, and orchestrating requests with a custom prompt pipeline. If you want to go further, you could even ensemble multiple LLMs for higher evenness accuracy.
2. It doesn't actually replace the modulo operator. Most people I know just use `n % 2 === 0` to check if a number is even, and they still keep that knowledge handy in case the AI service is down. This does not solve the reliability issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the library, is it reasonable to expect to make money off of this?
1. For an AI engineer, you can already build such a system yourself quite trivially by fine-tuning a lightweight inference model, deploying it behind a FastAPI endpoint, and orchestrating requests with a custom prompt pipeline. If you want to go further, you could even ensemble multiple LLMs for higher evenness accuracy.
2. It doesn't actually replace the modulo operator. Most people I know just use `n % 2 === 0` to check if a number is even, and they still keep that knowledge handy in case the AI service is down. This does not solve the reliability issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the library, is it reasonable to expect to make money off of this?