Dear #academia: Does anyone have any suggestions that will convert a PDF of an academic paper to an audio overview (say, 10-15 minutes) that is suitable for listening while driving
-
Dear #academia:
Does anyone have any suggestions that will convert a PDF of an academic paper to an audio overview (say, 10-15 minutes) that is suitable for listening while driving?
I've been using #notebooklm, but (1) I find the style of over-enthusiastic back-and-forth discourse a bit annoying, (2) it tends to oversimplify, and (3) it's too sycophantic -- everything about every paper is apparently really great and innovative.
At the other extreme, I don't want the paper read to me verbatim -- that's too much for me to process while driving. My high-level goal is to get the gist of a paper and to decide if it's worth my reading it.
-
Dear #academia:
Does anyone have any suggestions that will convert a PDF of an academic paper to an audio overview (say, 10-15 minutes) that is suitable for listening while driving?
I've been using #notebooklm, but (1) I find the style of over-enthusiastic back-and-forth discourse a bit annoying, (2) it tends to oversimplify, and (3) it's too sycophantic -- everything about every paper is apparently really great and innovative.
At the other extreme, I don't want the paper read to me verbatim -- that's too much for me to process while driving. My high-level goal is to get the gist of a paper and to decide if it's worth my reading it.
@msherr I don't know of such a thing (maybe it exists, dunno), but if it *does* exist and isn't called "Reviewer #2", that's a major missed opportunity.
-
@msherr I don't know of such a thing (maybe it exists, dunno), but if it *does* exist and isn't called "Reviewer #2", that's a major missed opportunity.
@mattblaze @msherr Nah—a Reviewer #2 bot would only say nasty things about the paper, while completely missing the point.
More seriously: I think what Micah is asking for is an AI-driven summary piped to a text-to-speech program. Given what I've seen of AI summarizers, I'm skeptical that such a program would be accurate enough to be useful, and I wonder how well text-to-speech would do on technical material. (A few weeks ago, Siri told me to take a road to somewhere named "Vah"—or at least that's how it decided to pronounce "VA." from a sign…) -
undefined Oblomov ha condiviso questa discussione