I had a stray idea that I should let all the local speech-to-text models battle it out over the same audio clip to figure out which ones to recommend and now, thanks to coding agents, I can just say it out loud and have it.
I had a stray idea that I should let all the local speech-to-text models battle it out over the same audio clip to figure out which ones to recommend and now, thanks to coding agents, I can just say it out loud and have it.