my basis is my gut. which is different from a code-based solution, in that it doesn’t care if it is factually wrong. the words are hitting the uncanny valley. it doesn’t matter if the LLM was used for content or translation.
as for the second question, it’s both. you can’t sell a model-based service as “as good as human” if the copy isn’t.
Thanks for the feedback. As noted, the website copy might need some work. We are a team of engineers, not copywriters, but we definitely have some work to do in that area.
Also considering this is a new product, we can only improve from here. Thanks.
Interesting, what's your basis for determining that an LLM was used for some of the copy?
> no, this is not enticing.
Can you please provide more context here? is this about the product itself or about an LLM being used for some of the copy?