Post #768242
2026-03-24 12:34 UTC
"In this work, we demonstrate that LLMs not only alter the voice and tone of human writing, but also consistently alter the intended meaning."
"heavy LLM users reported that the writing was less creative and not in their voice."
"Even when LLMs are prompted with expert feedback and asked to only make grammar edits, they still change the text in a way that significantly alters its semantic meaning."
"the LLM is not merely correcting grammar, but is actively steering diverse human perspectives towards homogenization, toward a different conceptual mode."
"extensive AI use results in a 70% change in the argumentative stance of essays, from for/against to neutral"
"LLMs systematically reframe arguments in more positive, optimistic terms, even when the original human text may have been critical or skeptical"
"LLMs have begun to change the very criteria that researchers use when evaluating peer-reviewed scientific research"
http://arxiv.org/abs/2603.18161
#llm #ai
Replies (2)
-
@ngaylinn@tech.lgbt 2026-03-24 21:08
I think what I find most upsetting about this is that the strongest bias they found in these models was towards ambivalence and positivity. That's because their weights are fine-tuned for agreeableness. It's also probably why people are satisfied with LLM-generated text, even though it isn't what they would have said or how they would have said it. But there's nothing special about agreeableness. There are a small number of people who could program whatever bias they like into those weights, and have it impressed upon our communication and thinking at a global scale.
-
@FrutigerAero00@mastodon.social 2026-03-25 15:25
@ngaylinn *pretends to be shocked *