Do AI Assistant "user guidelines" make it useless for attorneys doing criminal law?
The question in the subject is rhetorical, because as of now the answer is "yes." I'm an attorney. I was testing AI Assistant by asking questions regarding a brief in a real world criminal case. AI Assistant refused to respond to several questions, indicating "the document, the question, or the response" violated "User Guidelines." Seriously? Imagine: guns and violence in a brief involving a criminal case.
I thought attorneys were part of this product's target market. At this point, it's useless for any attorney doing criminal work. I can easily see it being useless for some species of civil work as well (for example, briefs involving allegations of s________l harassment).* Luckily, I have about a week left in which I can get a refund. I'm glad I caught it now rather than later.
*Ha! I received this warning when I tried to publish the post: "The message body contains s________l, which is not permitted in this community." So I had to obfuscate it like a scolded teenager. Grow up, Adobe.
