Built a skill to make Al avoid obvious Al-sounding phrases
heya heya, i made this skill: <in comments>
it makes Al writing less obviously Al-sounding. right now, it only counters the things mentioned in the research papers in the repo, so it definitely won't make Al write exactly like a human. that's kinda impossible anyway, since Al will always hallucinate at some point unless you feed it millions of tokens of context, which can just make lower-end models fuck up in other ways.