The only reasonable use of LLMs is when it’s fine if the answer is wrong. Those use cases exist, but apparently not what people typically do with them.
The only reasonable use of LLMs is when it’s fine if the answer is wrong. Those use cases exist, but apparently not what people typically do with them.
My one use case: helping practice an uncommon language I am learning. It doesn’t need to be correct about content, it just needs to respond plausibly. That’s exactly what they’re good at! I have yet to find a second use case.