A bill under consideration in New York would provide a private right of action, allowing people to file lawsuits against chatbot owners who violate the law.
I’m sure the quality of the LLM output does vary a lot based on the size of the scope it covers and the training data set.
However, I believe that if it were possible to get an LLM to be “quite accurate” in any context, that would make it easy to find a path to profitability for that tool, but I don’t think we have seen that materialize anywhere.
I believe that the best they can get is “more accurate” than the mean, but still not accurate enough to reliably make anyone money*.
It depends on the LLM actually.
Specialized medical LLMs are actually very accurate.
I’m sure the quality of the LLM output does vary a lot based on the size of the scope it covers and the training data set.
However, I believe that if it were possible to get an LLM to be “quite accurate” in any context, that would make it easy to find a path to profitability for that tool, but I don’t think we have seen that materialize anywhere.
I believe that the best they can get is “more accurate” than the mean, but still not accurate enough to reliably make anyone money*.
*Nvidia notwithstanding
Moreover, until you can get the same output from the same input from an LLM consistently, the entire tech is unreliable garbage.