This roughly checks out. I’m getting 66%, based on the methodology of cutting out the jug’s shape from the picture and numerically integrating the filled and empty volume (e.g. if a row is d pixels wide, it contributes d^2 to the volume, either filled or empty depending on whether it’s above or below the water level).

Note that openai’s original whisper models are pretty slow; in my experience the distil-whisper project (via a tool like whisperx) is more than 10x faster.

Really? This is the opposite of my experience with (distil-)whisper - I use it to generate subtitles for stuff like podcasts and was stunned at first by how high-quality the results are. I typically use distil-whisper/distil-large-v3, locally. Was it among the models you tried?

What’s so hilarious about it?

This post is currently top-1 when sorting by controversial. Objectively amazing bait.