|
|
|
|
|
by ithkuil
895 days ago
|
|
You may well be right about LLMs but if they can be adapted to match you're expectations about gender says something about how LLM works and not necessarily about how humans work My point is that the two things and not strictly related and you cannot make conclusions about humans just because you know something about LLMs or the converse. Since concepts like gender are to be understood within a context of humans I find sentences like "since gender is a social construct in humans then it follows that it must be a social construct in LLMs and necessarily easily changeable" to be fallacious logic |
|