I’d like to suggest a change (or perhaps optional setting) to the Explain screen. Right now it tends to give me two paragraphs: The description of the word’s meaning in English, and a few alternate English translations. Then there’s a whole second paragraph that tells me how it interprets the sentence the word came from. It’s pretty useless because it doesn’t seem to look at any context before or after the sentence; it’s basically just restating the translation. I never, ever read this paragraph and would like to be able to turn it off. What I would LOVE, however, is if the Explain screen offered me synonyms in my target language as well. I’m using the slightly rude screenshot below to make my point: It tells me this word is a rude version of “to pee”, which is good to know, but it would be super useful if at the end it also told me the actual normal word for “to pee”. Does that make sense?
To be honest I often appreciate the context explanations especially for complex sentences or metaphors as in your case. I probably wouldn’t have understood this metaphor without the explanation or I wouldn’t have been sure about it ![]()
But I would also appreciate a way to customize the explain prompt. It would be great if we could toggle parts like the context explanation or additional stuff like etymology, usage notes, synonyms, …
I see your point although in this case it’s not actually a metaphor!! A guy is describing the pain of having a urinary catheter ripped out too fast. Which is clear from the context before this sentence, but the AI doesn’t base its explanation on more than this one sentence, so it’s giving me this bogus metaphor analysis that actually has nothing to do with that text I’m reading. That’s what makes that second paragraph so pointless for me. Being able to customize the base prompt would be amazing; someone suggested that recently on another thread as well.
Ah, ok, I understand ![]()
Sometimes it’s funny though. I read Blood Weddings by Federico Garcia Lorca, which contains a lot of stuff referring to death and violence and the AI mostly tried to interpret it as a metaphor ![]()
Sometimes there might have been a double meaning, but I got the impression that when “in doubt” the AI preferred the metaphorical meaning over the literal one.
Yeah, I just don’t think the AI analysis is helpful if it’s not looking at more context than the individual sentence. But if some people find it useful, I’m not going to advocate for doing away it - just let us customize the prompt, Steve. ![]()
