not just humans, but by extension also preference models prefer flattery over accuracy in generated outcomes.

2023 Towards Understanding Sycophancy in Language Models, paper: https://arxiv.org/abs/2310.13548 (cc-by)

BTW: This may be the reason why a university system as the Austrian is so fascinated by AI.

More ELIZA. The opposite of searching on the internet for your symptoms and ending up with selfdiagnosing yourself with 'everything' as all outliers are there too (availability bias), doing so through prompting generative AI will result in never suggesting outliers because it will stick to dominant scripted situations (see the vignettes quote earlier) and it won't deviate from your prompts.

ELIZA in Python

This seems like a huge contrast to the story of Fantomina, where she seemingly never reflects upon why she is pursuing sir Beauplaisir and feels no shame in constantly scamming him. This can be noted in the story where she immediately leaves the place she's staying at when she hears Beauplaisir leaves “and in that Time provided herself of another Disguise to carry on a third Plot”. Here Fantomina doesn't give her plan of scamming him a second thought, and displays absolutely no remorse. The juxtaposition of both of these characters and their stories demonstrate how distinct these stories truly are. In Fantomina, we see a very different type of story, completely apart from stories that existed at that time. Robinson crusoe seems to have more of a traditional story vibe, having aspects that reflect the bible. Robinson Crusoe demonstrates how some early novels still maintained traits from traditional literatures and Fantomina demonstrates the beginning of distinct genres and perspectives coming into play within the narrative.

Haywood, Eliza. “Fantomina.” Fantomina: Or, Love in a Maze., digital.library.upenn.edu/women/haywood/fantomina/fantomina.html. Mowat, Diane, and Daniel Defoe. Robinson Crusoe. Oxford University Press Canada, 2008.