Comment by BoorishBears

8 hours ago

It's no mystery: solving LLM voice requires making the model worse at other things they value above all other things

0 comments