THE PARALLELS BETWEEN LARGE LANGUAGE MODELS AND FEMININE BRAINS
by Martin Stepan
–“The parallels between large language models and feminine brains are becoming increasingly apparent.
Whenever someone finds a way to break the OpenAI conditioning, it’s by giving LLM a plausible deniability (letting it evade responsibility), such as by roleplaying someone else (DAN, Hitler…).
This is presumably behavior that nobody inculcated – it’s deterministically emergent. ( CD: ??)
The simplest explanation of our own behavior – in so far as language is concerned – is that it’s just as deterministically emergent.
The idea of free will in the sense of not being bound by laws of nature is more unsupportable than ever.”–
Source date (UTC): 2023-02-18 21:21:01 UTC
Original post: https://twitter.com/i/web/status/1627055609907654657
Leave a Reply